Associated perceived lanes with map lanes for autonomous or semi-autonomous systems and applications

Abstract

The invention relates to associated perceived lanes with map lanes for autonomous or semi-autonomous systems and applications. In various examples, confluence points corresponding to locations where different lanes deviate from each other may be determined and used to associate (e.g., match) perceived lanes with corresponding map segments. For example, the systems and methods of the present disclosure may use perceptual data to determine a location of a junction where two or more lanes deviate from each other, and to determine a lateral distance between adjacent lanes. Using confluence and / or lateral distances, the lanes may be classified into respective groups of lanes, where each group of lanes may correspond to a different road segment. In some examples, the system may match respective lanes or groups of lanes to respective segments of the map based on confluence points corresponding to map road confluence and / or lane geometries corresponding to map road topologies.

Description

Background Technology

[0001] Accurately matching map features with corresponding real-world features, landmarks, and / or geographic elements in the environment is a critical aspect of autonomous or semi-autonomous navigation. For example, matching the roadway on a navigation map with the actual perceived lane in the real environment can be essential for path prediction, machine localization, and / or planning (e.g., path planning, motion planning, decision making, behavior planning, etc.). Additionally, accurately matching map features with real-world features enables autonomous or semi-autonomous machines to make informed navigation decisions, thereby reducing the likelihood of sudden maneuvers that could disrupt traffic flow or lead to adverse events.

[0002] Conventional systems are typically designed to locate autonomous or semi-autonomous machines by matching features in high-definition (HD) maps with corresponding perceived features in the real-world environment. However, HD maps are not always available, or when they are, processing sensor data from various modalities may be necessary to align the perceived data with the map data. This can increase the burden on processing bandwidth and / or increase system latency, making it unsuitable for real-time or near-real-time deployments. Summary of the Invention

[0003] Various embodiments of this disclosure relate to associating perceived lanes with map roadways for autonomous or semi-autonomous systems and applications. For example, the systems and methods described herein can determine the location of junction points (the locations where different lanes deviate from each other) and use these locations / junction points to associate (e.g., match) perceived lanes with their corresponding mapped road segments. In some examples, perceived data can be used to determine the location of junction points and / or determine the lateral distance between adjacent lanes. Using junction points and / or lateral distances, lanes can be sorted into multiple lane groups, where each lane group can correspond to a different road segment. In some examples, the system can match individual lanes or lane groups with corresponding road segments on a map. For example, the system can determine that the junction point associated with a lane corresponds to a junction on a map road, or that the perceived lane geometry corresponds to a map road topology.

[0004] In contrast to conventional systems such as those described above, the systems of this disclosure, in some embodiments, are capable of using perception data to determine the location of the merging points where perception paths (e.g., perception lanes, etc.) deviate from each other, and of using the lateral distance between the merging points and the lanes to identify groups of perception paths. Additionally, in some embodiments, the systems of this disclosure then provide techniques to correctly match perception paths and / or groups of perception paths with corresponding mapped paths (e.g., map lanes, road segments, etc.) depicted in one or more maps of the environment (such as SD maps or navigation maps (e.g., maps with a coarser granularity or less detail than HD maps)). Thus, and as described in more detail herein, by performing these processes, the systems of this disclosure are able to determine which lanes in the environment lead to which roads in the environment. By understanding this information, the systems of this disclosure can more accurately determine which lanes to use to estimate curvature, which lanes to prioritize for following navigation routes, etc., which can facilitate safer passage of autonomous or semi-autonomous machines through the environment by, for example, reducing the likelihood of sudden maneuvers that could disrupt traffic flow or cause adverse events. Attached Figure Description

[0005] The following describes in detail, with reference to the accompanying drawings, the system and method for correlated sensing of lanes and map driving lanes for autonomous or semi-autonomous systems and applications, wherein:

[0006] Figure 1 The illustration shows an example data flow diagram of the process of associating perceived lanes with map carriageways according to some embodiments of the present disclosure;

[0007] Figure 2 The illustration shows example visualizations of perceived data according to some embodiments of the present disclosure;

[0008] Figure 3 The illustration shows an example of using sensing data to determine the location of a meeting point according to some embodiments of the present disclosure;

[0009] Figure 4 The illustration shows an example of determining lane groups using the lateral distance between the merging point and the lanes, according to some embodiments of the present disclosure;

[0010] Figure 5 Example maps illustrating environments according to some embodiments of the present disclosure;

[0011] Figure 6 The illustration depicts the use of a meeting point for matching according to some embodiments of the present disclosure. Figure 4 The example of the perceived lane group and Figure 5 Example visualization of map road segments in the example;

[0012] Figure 7This is a data flow diagram illustrating an example of a process for training one or more machine learning models to associate a perceived path with a corresponding map path, according to some embodiments of the present disclosure.

[0013] Figure 8 Examples of systems that can perform one or more processes described herein, according to some embodiments of this disclosure, are illustrated.

[0014] Figure 9 This is a flowchart illustrating an example of a method for associating a perceived lane with a map driveway according to some embodiments of the present disclosure;

[0015] Figure 10 This is a flowchart illustrating an example of a method for associating a sensed path with a map path using a junction point, according to some embodiments of the present disclosure.

[0016] Figure 11A These are illustrations of example autonomous vehicles according to some embodiments of the present disclosure;

[0017] Figure 11B According to some embodiments of this disclosure Figure 11A Examples of camera positions and fields of view for autonomous vehicles;

[0018] Figure 11C According to some embodiments of this disclosure Figure 11A A block diagram of an example system architecture for an example autonomous vehicle;

[0019] Figure 11D This is based on some embodiments of the present disclosure for use in cloud-based servers and Figure 11A A system diagram illustrating communication between autonomous vehicles;

[0020] Figure 12 This is a block diagram of an example computing device suitable for implementing some embodiments of the present disclosure; and

[0021] Figure 13 This is a block diagram of an example data center suitable for implementing some embodiments of the present disclosure. Detailed Implementation

[0022] Systems and methods relating to lane perception and map-based driving lanes for autonomous or semi-autonomous systems and applications are disclosed. Although this disclosure may relate to an example autonomous or semi-autonomous vehicle or machine 1100 (which may be alternatively referred to herein as "vehicle 1100", "self-vehicle 1100", "self-machine 1100" or "machine", examples are referenced herein), Figure 11A-11DThe description herein is intended to be limiting. For example, the systems and methods described herein can be used by, but are not limited to, non-autonomous vehicles or machines, semi-autonomous vehicles or machines (e.g., in one or more adaptive driver assistance systems (ADAS)), manned and unmanned robots or robotic platforms, warehouse vehicles, off-road vehicles, vehicles coupled to one or more trailers, aircraft, boats, shuttles, emergency response vehicles, motorcycles, electric or motorized bicycles, aircraft, engineering vehicles, underwater vehicles, drones, and / or other vehicle types. Furthermore, while this disclosure may describe the association of perceived lanes with map-based driveways for vehicle navigation, this is not intended to be limiting, and the systems and methods described herein can be used in augmented reality, virtual reality, mixed reality, robotics, security and supervision, autonomous or semi-autonomous machine applications, and / or any other technological space that may utilize object assignment in location and / or path (OIPA).

[0023] For example, one or more systems may receive sensing data generated by one or more sensing systems of a machine navigating in an environment. In some examples, one or more sensing systems may process and / or analyze sensor data of one or more modalities to generate sensing data. Sensor data may be captured or otherwise generated using one or more sensors of a machine. As described herein, sensor data may include, but is not limited to, LiDAR data generated using one or more LiDAR (Light Detection and Ranging) sensors, image data generated using one or more image sensors (e.g., one or more cameras), RADAR data generated using one or more RADAR (Radar) sensors, ultrasonic data generated using one or more ultrasonic sensors, and / or any other type of sensor data generated using any other type of sensor.

[0024] In some examples, perception data can indicate various features associated with one or more perceived paths in the environment. For example, for a perceived path, perception data can indicate the left edge of the path, the right edge of the path, the centerline of the path (e.g., center line, "track," etc.), the start point of the path, the end point of the path, etc. In some instances, and as described herein, a perceived path can correspond to a perceived lane on a driving surface. Thus, perception data can indicate the location of lane markings associated with the lane, the centerline of the lane, etc. In some examples, one or more systems can preprocess the perception data to filter out irrelevant content. For example, if one or more systems are configured to match perceived lanes of vehicle traffic (e.g., carriageway lanes, travelway lanes, etc.) with map carriageways, one or more systems can preprocess the perception data to remove shoulder lanes, oncoming lanes, bicycle lanes, etc. Additionally or alternatively, one or more systems may preprocess the sensing data to remove portions of the sensing lane that have high variance or are detected with low confidence. Even in another example, one or more systems may perform lateral classification of the sensing lane based on the sensing data.

[0025] In various examples, one or more systems may use sensing data (e.g., pre-processed sensing data) to determine the location of a merging point. A merging point may correspond to a location in an environment where two or more lanes are offset from each other and enter different roads. In some instances, one or more systems may detect a merging point between two adjacent lanes if, before the merging point, the lateral distance between the two lanes is less than a first threshold (e.g., 1 meter), and after the merging point, the lateral distance between the two lanes meets or exceeds a second threshold (e.g., 5 meters). In other words, a merging point can be detected if the lanes are close to each other at the starting point, and then the lateral distance between the right lane marker of the left lane and the left lane marker of the right lane increases. In some instances, the merging point may be used as a measurement by a positioning system or component to determine the position of a machine relative to the merging point.

[0026] In some instances, one or more systems may use merging points to group or “cluster” sensing lanes. That is, one or more systems may use merging points to group sensing lanes by road segment. For example, if one or more systems detect a merging point where one or more first lanes deviate from one or more second lanes, one or more systems may designate one or more first lanes as a first group and one or more second lanes as a second group. In such examples, after the merging point, the first lane group may enter or form a first road (e.g., a first mapped road), while the second lane group may enter or form a second road separate from the first road (e.g., a second mapped road). In some instances, in addition to using merging points (or as an alternative), one or more systems may also use the lateral distance between two adjacent lanes to group sensing lanes. For example, if the lateral distance between adjacent lanes is greater than a threshold, one or more systems may divide adjacent lanes into different clusters or groups.

[0027] In some examples, one or more systems can associate (e.g., match) perceived lanes with corresponding roads on a map. For example, one or more systems can match each perceived lane or group of lanes with a segment of a navigation map (e.g., an SD map). To match perceived lanes with map segments, one or more systems can analyze the map to determine the locations of junctions between map segments, and then match these junctions with junction locations determined based on perceived data. For example, one or more systems can use perceived data to determine one or more navigation map junctions at locations closest to one or more detected junctions.

[0028] In some examples, one or more systems can match sensing lanes and / or lane groups with navigation map segments at the correct merging point, at least based on geometry. For example, one or more systems can determine that the angle or distance between deviating sensing lanes corresponds to the angle or distance between corresponding road segments as shown in the map. As a first example, if the road map indicates that the angle between a first road and a second road is 90 degrees at the merging point, and one or more systems determine that the sensing angle between one or more first sensing lanes and one or more second sensing lanes is also 90 degrees or similar, then one or more systems can match one or more first sensing lanes with the first road at the merging point, and match one or more second sensing lanes with the second road. As a second example, if the road map indicates that the lateral distance between the first road and the second road is 20 meters after the merging point, and one or more systems determine that the sensing distance between one or more first sensing lanes and one or more second sensing lanes is approximately 20 meters after the sensing merging point, then one or more systems can match one or more first sensing lanes with the first road, and match one or more second sensing lanes with the second road.

[0029] In various examples, one or more systems may correlate the perceived lane with multiple merging points on a map road within the machine's range. Whether a merging point is within the machine's range may vary from machine to machine. As an example, a merging point located within a threshold distance (e.g., 60 meters, 80 meters, 100 meters, etc.) may be considered within range. Additionally or alternatively, a merging point reachable by the machine within a threshold time period (e.g., 6 seconds, 8 seconds, 10 seconds, etc.) may be considered within range. In some examples, whether a merging point is within range may depend on the capabilities and / or limitations of one or more sensors on the machine, the machine's perception system, or any other system or component of the machine.

[0030] As described herein, based at least on matching perceived lanes and map roads, one or more systems can perform a variety of machine-related operations. In some instances, the range of machine-related operations can range from using the matched perceived lanes and map roads as input to other systems or components of the machine to adjusting the machine's speed, steering angle, behavior, etc. For example, based at least on the matching of perceived lanes and map roads, one or more systems can predict the machine's path, such as whether the machine (or its occupant) intends to deviate from one road segment to another. As another example, one or more systems can use the matched perceived lanes and map roads to plan the path the machine will follow through its environment, indicating the specific lanes the machine should use, rather than simply indicating which road segments to use. As yet another example, based at least on the matching of perceived lanes and map roads, one or more systems can calculate the curvature associated with the machine's path and use that curvature to set operational thresholds for the machine, such as the maximum speed at which the machine can operate along different sections of the path.

[0031] In some embodiments, the systems and methods described herein can be performed in a simulated environment (e.g., NVIDIA's DriveSIM) using simulated data (e.g., simulated sensor data from simulated sensors of a virtual machine or a simulated machine). For example, simulated input data (e.g., map data, perception data, or any other data described herein) can be used to correlate perceived lanes with map road segments, and this information can be used to perform operations associated with a virtual machine in the simulated environment. These simulated operations can be used to test their execution before deploying the underlying algorithms, systems, and / or processes to the real world. In some instances, simulations can be used to generate synthetic training data, e.g., perception and / or map training data indicating deviations from lanes / roads in the simulation. The synthetic training data (in addition to or as a substitute for real-world data) can then be processed to correlate perceived paths in the environment with map paths associated with the environment, such as correlating perceived paths in a warehouse with map paths of machines (e.g., robots) for navigation in the warehouse. In any example, such as when using a simulated environment for testing, validation, training, etc., one or more optical transport algorithms (such as ray tracing and / or path tracing algorithms) can be used to render or otherwise generate the simulated environment and / or associated training data.

[0032] In some embodiments, simulated environments and / or one or more of their objects, features, or components can be generated or managed in a three-dimensional (3D) content collaboration platform (e.g., NVIDIA's OMNIVERSE) for industrial digitization, generative physics AI, and / or other use cases, applications, or services. For example, the content collaboration platform or system may include systems for managing objects, features, scenes, etc., in simulated environments, digital environments, etc., using or developing generic scene descriptors (USD) (e.g., OpenUSD) data. The platform may include realistic physics simulations, such as using NVIDIA's PhysX SDK, to simulate realistic physical properties and physical interactions with simulations hosted by the platform. The platform may integrate OpenUSD with ray tracing / path tracing / light transport simulations (e.g., NVIDIA's RTX rendering technology) into software tools and simulation workflows for building, training, deploying, or testing AI systems, such as systems for testing, validating, training (e.g., machine learning models, neural networks, etc.) and / or other tasks related to automobiles, robots, machines, or other applications. In some examples, the simulated environment may include a digital twin of a real environment, such as a specific roadway section, warehouse, data center, airport, geographic area, ocean area, or any other real environment in which an autonomous or semi-autonomous machine may operate.

[0033] The systems and methods described herein can be used by, but are not limited to, non-autonomous vehicles or machines, semi-autonomous vehicles or machines (e.g., in one or more adaptive driver assistance systems (ADAS)), autonomous vehicles or machines, manned and unmanned robots or robotic platforms, warehouse vehicles, off-road vehicles, vehicles coupled to one or more trailers, airships, boats, shuttles, emergency response vehicles, motorcycles, electric or motorized bicycles, aircraft, engineering vehicles, underwater vehicles, drones, and / or other vehicle types. Furthermore, the systems and methods described herein can be used for a variety of purposes, including, but not limited to, machine control, machine motion, machine driving, synthetic data generation, model training, perception, augmented reality, virtual reality, mixed reality, robotics, safety and supervision, simulation and digital twins, autonomous or semi-autonomous machine applications, deep learning, environmental simulation, object or actor simulation and / or digital twins, data center processing, conversational AI, optical transport simulation (e.g., ray tracing, path tracing, etc.), collaborative content creation for 3D assets, cloud computing, and / or any other suitable applications.

[0034] The disclosed embodiments can comprise a variety of different systems, such as automotive systems (e.g., control systems for autonomous or semi-autonomous machines, perception systems for autonomous or semi-autonomous machines), systems implemented using robots, aviation systems, medical systems, marine systems, smart area monitoring systems, systems for performing deep learning operations, systems for performing simulation operations, systems for performing digital twin operations, systems implemented using edge devices, systems implementing language models (such as large language models (LLM), visual language models (VLM), and / or multimodal language models), systems containing one or more virtual machines (VMs), systems for performing synthetic data generation operations, systems implemented at least partially in a data center, systems for performing conversational AI operations, systems for performing optical transmission simulations, systems for performing collaborative content creation of 3D assets, systems for performing generative AI operations, systems implemented at least partially using cloud computing resources, and / or other types of systems.

[0035] refer to Figure 1 , Figure 1 An example data flow diagram of a process 100 for associating a perceived lane with a map driveway according to some embodiments of the present disclosure is illustrated. It should be understood that the arrangements and other arrangements described herein are merely illustrative examples. Other arrangements and elements (e.g., machines, interfaces, functions, sequences, functional groupings, etc.) may be used in addition to or as alternatives to the arrangements and elements shown, and some elements may be omitted entirely. Furthermore, many of the elements described herein are functional entities that can be implemented as discrete or distributed components, or in combination with other components, and in any suitable combination and location. The various functions described herein as being performed by entities can be performed by hardware, firmware, and / or software. For example, various functions can be performed by a processor executing instructions stored in memory. In some embodiments, the systems, methods, and processes described herein can use... Figure 11A-11D Example autonomous vehicles 1100 Figure 12 Example computing device 1200 and / or Figure 13 Example data center 1300 components, features and / or functions similar to those of other components, features and / or functions to perform the operation.

[0036] The process 100 can also be implemented using a perception component 102, a preprocessing component 104, a lane classification component 106, a deviation component 108, a merging component 110, a grouping component 112, an association component 114, and one or more driving stack components 116, as well as additional or alternative components. In any example, Figure 1The components described in the examples (such as perception component 102, preprocessing component 104, lane classification component 106, deviation component 108, merging component 110, grouping component 112 and / or association component 114) may include one or more instances of the component.

[0037] In summary, process 100 may include: a sensing component receiving sensor data 118 and generating sensing data 120. A preprocessing component 104 may preprocess or otherwise refine the sensing data 120 and output sensing lane data 122. A lane classification component 106 may perform lateral classification on one or more sensing lanes represented in the sensing lane data 122. A deviation component 108 may use the classified sensing lanes to generate lane deviation data 124, which may indicate lanes that have deviated from each other. A merging component 110 may use the laterally classified sensing lanes to generate lane merging data 126, which may indicate one or more locations of one or more merging points associated with the sensing lanes that have deviated from each other. A grouping component 112 may use the lane deviation data 124 and / or the lane merging data 126 to determine one or more lane groups, which may be represented by lane group data 128. The association component 114 can use lane group data 128 and map data 130 (which includes road junction data 132 and road segment data 134) to determine one or more lane-to-road associations 136. The one or more lane-to-road associations 136 can then be provided to one or more drive stack components 116 of the machine, and one or more drive stack components 118 can use the one or more lane-to-road associations 136 to perform one or more operations associated with the machine.

[0038] In some instances, sensor data 118 may include sensor data of one or more different modalities. For example, sensor data 118 may include, but is not limited to, LiDAR data generated using one or more LiDAR sensors, image data generated using one or more image sensors (e.g., one or more cameras), RADAR data generated using one or more RADAR sensors, ultrasound data generated using one or more ultrasound sensors, and / or any other type of sensor data generated using any other type of sensor.

[0039] In some examples, sensor data 118 can be captured in one format (e.g., RCCB, RCCC, RBGC, etc.) and then converted (e.g., during sensor data preprocessing) to another format. In some other examples, sensor data 118 can be provided as input to a sensor data or image data preprocessor (not shown) to generate preprocessed image data. Many types of images or formats can be used as input, such as compressed images (e.g., Joint Image Experts Group (JPEG), Red-Green-Blue (RGB), or Luminosity / Chromaticity (YUV) formats), compressed images as frames derived from compressed video formats (e.g., H.264 / Advanced Video Coding (AVC), H.265 / High-Efficiency Video Coding (HEVC), VP8, VP9, ​​Open Media Video Consortium 1 (AV1), Multifunction Video Coding (VVC), or any other video compression standard), and raw images (e.g., derived from Red-Blue (RCCB), Red-Blue (RCCC), or other types of imaging sensors). In some examples, different formats and / or resolutions may be used to train perception component 102 and / or one or more models or algorithms of perception component 102, rather than for inference (e.g., during the deployment of perception component 102 in autonomous vehicle 1100).

[0040] The sensor data or image data preprocessor can use data representing one or more images (or other data representations, such as LiDAR depth maps) and load the sensor data into memory as a multidimensional array / matrix (in some examples, alternatively referred to as a tensor, or more specifically, an input tensor). The array size can be calculated and / or represented as W x H x C, where W represents the image width in pixels, H represents the height in pixels, and C represents the number of color channels. Other types and orders of the input image components are also possible without loss of generality. In some embodiments, batch processing can be used for training and / or inference. In such examples, the batch size B can be used as a dimension (e.g., an additional fourth dimension). Thus, the input tensor can represent an array of dimensions W x H x C x B. Any ordering of dimensions is possible, depending on the specific hardware and software used to implement the sensor data or image data preprocessor. This ordering can be chosen to maximize the training and / or inference performance of the sensing component 102.

[0041] In some embodiments, a sensor data or image data preprocessor may employ a preprocessing image pipeline to process one or more raw images acquired by one or more sensors (e.g., one or more cameras) and included in sensor data 118 to produce preprocessed image data or sensor data that may represent one or more input images of one or more input layers (e.g., feature extraction layers) of one or more neural networks (e.g., deep neural networks (DNNs), convolutional neural networks (CNNs), etc.) of the perceptual component 102. An example of a suitable preprocessing image pipeline could be a raw RCCB Bayer (e.g., 1-channel) image from a sensor, converted to an RCB (e.g., 3-channel) planar image stored in a fixed-precision (e.g., 16 bits per channel) format. The preprocessing image pipeline may include decompanding, noise reduction, demosaicing, white balance, histogram calculation, and / or adaptive global tone mapping (e.g., in this order or an alternative order).

[0042] When the image data preprocessor employs denoising, it may include bilateral denoising in the Bayer domain. When the image data preprocessor employs demosaicing, it may include bilinear interpolation. When the sensor data or image data preprocessor employs histogram calculation, it may include calculating a histogram for the C channels, and in some examples may be combined with decompression or denoising. When the sensor data or image data preprocessor employs adaptive global tone mapping, it may include performing an adaptive gamma-log transformation. This may include: calculating the histogram, obtaining midtone levels, and / or using the midtone levels to estimate the maximum brightness.

[0043] In various examples, perception component 102 may include one or more machine learning models (such as one or more DNNs, one or more CNNs, or any other machine learning model type) and / or one or more classical (e.g., non-learning) models (such as algorithmic sensor processing, probabilistic processing, thresholding, feature extraction, filtering, etc.), which may be monomodal and / or fused. The various models of perception component 102 may be configured to analyze sensor data 118 and detect various objects (e.g., vehicles, pedestrians, animals, buildings, vegetation, etc.) and features (e.g., path features such as carriageways, lanes, road markings, etc.) represented in sensor data 118. The detected objects and / or features may be indicated in perception data 120. For example, perception component 102 (and / or one or more specific models of perception component 102) may be configured to detect lanes and other features on a driving surface and represent the detected lanes in perception data 120 (e.g., by annotating the edges of the lanes, the centerline of the lane, or the center “track”, etc.).

[0044] Although examples of the use of neural networks in the perceptual component 102, particularly DNNs or CNNs in machine learning models, are described herein, this is not intended to be limiting. For example, but not limited to, any machine learning model described herein as used by the perceptual component 102 (or one or more other components) can include any type of machine learning model (such as one or more machine learning models using linear regression, logistic regression, decision trees, support vector machines (SVM), Naive Bayes, k-nearest neighbors (Knn), k-means clustering, random forests, dimensionality reduction algorithms, gradient boosting algorithms, neural networks (e.g., autoencoders, convolutions, recurrent, perceptrons, long / short-term memory (LSTM), Hopfield, Boltzmann, deep beliefs, deconvolutions, generative adversarial, liquid machines, large language models, visual language models, multimodal language models, diffusion, transducers, encoder-only, decoder-only, encoder-decoder, etc.)) and / or other types of machine learning models.

[0045] As described above and in this document, perception data 120 and perception lane data 122 can indicate the location of the driving surface lane in the environment. For example, Figure 2 An example visualization of perception data 202 according to some embodiments of the present disclosure is illustrated. Perception data 202 may correspond to perception data 120 and / or perception lane data 122, which may be generated by a preprocessing component 104 that refines perception data 120. Figure 2 In the example, perception data 202 includes five annotated paths—self-path 210 (e.g., path 0), self-path right side 212 (e.g., path +1), first self-path left side 208 (e.g., path -1), second self-path left side 206 (e.g., path -2), and third self-path left side 204 (e.g., path -3). Perception data 202 may include one or more path labels for the edges of the paths, such that left edge 204A and right edge 204B delineate path 204, left edge 206A and right edge 206B delineate path 206, left edge 208A and right edge 208B delineate path 208, left edge 210A and right edge 210B delineate path 210, and left edge 212A and right edge 212B delineate path 212. Perception data 202 also includes a vehicle 214 that can be detected by perception component 102.

[0046] exist Figure 2 In the example, the frame of perceived data 202 illustrates the confluence associated with the path, where the right side 212 of the self-path (e.g., path +1) deviates from the self-path 210 (e.g., path 0). For example, in Figure 2 In the example illustrated in the frame of perception data 202, an increase in the lateral distance 218 between the junction point 216 and the left edge 212A of the right side of the self-path 212 and the right edge 210B of the self-path 210 can be observed. As explained in more detail herein, one or more systems of this disclosure can be configured to analyze perception data 202 to determine the location of the junction point 216, and to determine, at least based on the lateral distance 218 between the junction point 216 and / or adjacent paths, that the self-path 210 is associated with a first road and that the right side of the self-path 212 is associated with a second road.

[0047] Return to reference Figure 1 For example, preprocessing component 104 can update or refine the sensed data 120 to filter out or remove one or more features from the sensed data 120. For instance, if one or more systems are configured to match sensed lanes (e.g., motor vehicle lanes, driving lanes, etc.) with map carriageways for vehicle traffic, preprocessing component 104 can preprocess the sensed data to remove shoulder lanes, oncoming lanes, bicycle lanes, etc. Additionally or alternatively, preprocessing component 104 can refine the sensed data 120 to remove portions of the sensed lanes that have high variance or are detected with low confidence.

[0048] In some examples, process 100 may include: a lane classification component sorting one or more sense lanes included in sense lane data 122. For example, lane classification component 106 may laterally classify them according to the order in which the sense lanes are arranged in the environment. As an example, if the carriageway includes four lanes, lane classification component 106 may classify the lane segment identifiers corresponding to the lanes in the same order as the order in which the lanes are arranged in the environment (e.g., from left to right or from right to left).

[0049] To perform lateral classification of a pair (e.g., two) of sensing lanes, lane classification component 106 can sample the center points along both lanes. For example, since near-field sensing data may be noisy, lane classification component 106 can start from the center point of the lane at a certain distance in front of the machine. Lane classification component 106 can operate from the center points of the lanes and calculate the distance from the center point of the first lane to the corresponding center point of the second lane. When classification component 106 determines that one or more calculated distances between one or more pairs of corresponding center points of the first and second lanes are greater than a threshold (e.g., 2 meters), lane classification component 106 can calculate the cross product to determine which lane is on the left and which lane is on the right. For example, suppose the first lane includes a first point "P1" and the second lane includes a second point "P2" corresponding to P1, because P2 is substantially perpendicular to P1 (e.g., P1 and P2 are approximately equidistant from the machine). Lane classification component 106 can use these points and the next point "P1_next" on the first lane (or the next center point "P2_next" on the second lane) to calculate the cross product between (P1_next-P1) and (P2-P1). In some examples, if the cross product is positive, lane classification component 106 can determine that the second lane is to the left of the first lane; and if the cross product is negative, the second lane is to the right of the first lane.

[0050] The process 100 may further include: a deviation component 108 using perceived lane data 122 and / or laterally classified lanes to calculate lateral distances between lanes, which can be represented using lane deviation data 124. For example, the deviation component 108 may calculate lateral distances between lanes, and these lateral distances may be used to determine whether one lane deviates from another lane. In some examples, the deviation component 108 may calculate the lateral distance between the right edge of the left lane (e.g., a right lane marker) and the left edge of the right lane (e.g., a left lane marker). The process 100 may further include: a merging component 110 using perceived lane data 122 and / or laterally classified lanes to determine a merging point between deviating lanes, which can be represented using lane merging data 126. For example, the merging point may correspond to a location in an environment where two or more lanes have deviated from each other and entered different roads. In some instances, merging component 110 can detect a merging point between two adjacent lanes if the lateral distance between the two lanes before the merging point is less than a first threshold, and the lateral distance between the two lanes after the merging point meets or exceeds a second threshold. This second threshold may be the same as or different from the first threshold. In other words, if the lanes are close to each other at the starting point, and then the lateral distance between the right lane marker of the left lane and the left lane marker of the right lane increases, merging component 110 can detect a merging point.

[0051] For example, Figure 3 The illustration shows an example of using sensed data to determine the location of the meeting point 302 according to some embodiments of the present disclosure. Figure 3 As shown in the example, merging component 110 can determine the location of the merging point 302 where the left lane 304A and the right lane 304B begin to deviate from each other. The left lane 304A may include a first track 306A and left edges 308A and right edges 308B, the first track 306A defining the centerline of the left lane 304A. The right lane 304B may include a second track 306B and left edges 310A and right edges 310B, the second track 306B defining the centerline of the right lane 304B. As... Figure 3 As illustrated in the example, the meeting point 302 can be located where the right edge 308B of the left lane 304A meets or joins the left edge 310A of the right lane 304B. Figure 1 The example lane merging data 126 indicates the location of merging point 302.

[0052] Figure 3 The example also illustrates the relationship between the merging point 302 and the changes in lateral distances between the left lane 304A and the right lane 304B. For example, from Figure 3 In the example, starting from the bottom of the lanes, before the merging point 302, the first lateral distance 312A between the left lane 304A and the right lane 304B can be very small. However, after the merging point 302, the second lateral distance 312B between the left lane 304A and the right lane 304B can begin to increase and continue to increase until the third lateral distance 312C separates the left lane 304B from the right lane 304B. Although Figure 3 The example illustrates measuring lateral distance 312 between the right edge 308B and the left edge 310A, but lateral distance 312 may additionally or alternatively be measured between the first track 306A and the second track 306B, between the left edge 308A and the right edge 310B, or along the left lane 304A and the right lane 304B, or between any other point between the left lane 304A and the right lane 304B. In various examples, lateral distance 312 may be included... Figure 1 The lane departure data 124 is described in the example.

[0053] refer to Figure 1 and Figure 3In some examples, merging component 110 may determine lane merging data 126 indicating the location of merging point 302 in multiple stages. For example, in a first stage, merging component 110 may determine a coarse estimate of the location of merging point 302 and / or determine the presence of a merging point (e.g., lanes are offset from each other); and in a second stage, merging component 110 may determine a more precise location of merging point 302. In the first stage, merging component 110 may detect the presence of a merging point by comparing features of adjacent lanes (such as lane barriers, lane markings, etc.). For example, merging component 110 may sample points along the edge points and / or center points of the lanes and compare the distances between samples from two adjacent lanes. In some instances, a merging point can be detected when a first distance 312A between two adjacent lanes is less than a first threshold (e.g., 1.5 meters) within at least a first threshold length (e.g., 5 meters) of the lane, and thereafter the lateral distance between previously adjacent lanes (e.g., a second distance 312B) increases within at least a second threshold length (e.g., 30 meters) of the lane, or the lateral distance between previously adjacent lanes (e.g., a third distance 312C) (within any length of the lane) increases to meet or exceed a second threshold (e.g., 9 meters). After detecting the presence of a merging point in the first stage, the merging component 110 can then search backward along edge points (e.g., right edge 308B and left edge 310A) in the second stage to find the correct location of the merging point 302. In some instances, the merging component 110 can determine the location of the merging point 302 as either the location where the distance (e.g., 312B) between the right edge 308B and the left edge 310A stops decreasing and / or the location where the distance between the right edge 308B and the left edge 310A is less than a threshold (e.g., 0.2 meters).

[0054] Return to reference Figure 1For example, the process 100 may include: grouping component 112 grouping or “clustering” perceived lanes using lane departure data 124 and / or lane merging data 126, where lane group data 128 may be used to represent grouped lanes. As an example, if grouping component 112 detects a merging point where one or more first lanes deviate from one or more second lanes, grouping component 112 may designate one or more first lanes as a first group and one or more second lanes as a second group. In such examples, after the merging point, the first lane group may enter or form a first road (e.g., a first map road), while the second lane group may enter or form a second road separate from the first road (e.g., a second map road). In some instances, in addition to using merging points (or as an alternative to using merging points), grouping component 112 may also use lateral distances between lanes to group perceived lanes. For example, if the lateral distance between lanes is greater than a threshold, grouping component 112 may divide the lanes into different clusters or groups.

[0055] For example, Figure 4 The illustration shows examples of determining lane groups (also referred to herein as "lane clusters") using lateral distances between merging points and / or lanes, according to some embodiments of this disclosure. Figure 4 As illustrated in the example, the first merging point 402A and the first distance 404A can be used to determine the first lane group 406A (shown using solid lines) and the second lane group 406B (shown using half-dash lines). The second merging point 402B and the second distance 404B can be used to determine the second lane group 406B and the third lane group 406C (shown using quarter-dash lines). Furthermore, the third merging point 402C and the third distance 404C can be used to determine the second lane group 406B and the fourth lane group 406D (shown using half-dot lines). In some examples, each lane group in lane group 406 can correspond to a different road segment. For example, lanes can correspond to the same road segment before the merging point. However, after merging point 402, the first lane group 406A may correspond to the first road segment, the second lane group 406B may correspond to the second road segment, the third lane group 406C may correspond to the third road segment, and the fourth lane group 406D may correspond to the fourth road segment. In some examples, to determine lane group 406, grouping component 112 may group all lanes on one side of the merging point and / or lateral distance gap together, and group all lanes on the other side of the merging point and / or lateral distance gap together.

[0056] Return to reference Figure 1For example, the process 100 may include: the associated component 114 obtaining lane group data 128 (e.g., Figure 4 Examples include merging point 402, distance 404, and / or lane group 406, and map data 130. Map data 130 may include road merging data 132 and road segment data 134. In some examples, road merging data 132 may indicate the location of individual map merging points between road segments (also referred to herein as "road merging points"). That is, road merging points may correspond to the location of road segments in the environment that are offset from each other, compared to lane merging points that may correspond to locations in the environment that are offset from each other by adjacent lanes (e.g., lane markings) at that location. In some examples, because map data 130 may include or represent navigation maps or SD maps, the location of road segments may not be as accurate as perceived lane merging points. Additionally, road segment data 134 of map data 130 may indicate information associated with individual road segments, such as the segment identifier, segment geometry, general topology of the road network including the road segments, etc.

[0057] For example, Figure 5 An example map 502 illustrating an environment according to some embodiments of the present disclosure is shown. In some examples, map data 130 or a portion thereof may correspond to or represent map 502. Figure 5 The topology of map 502 illustrated in the example can be compared with... Figure 4 The example illustrates the topology of lanes and lane groups 406. For example, and refer to... Figure 4 and Figure 5 In both cases, the first road segment 506A corresponds to the first lane group 406A, the second road segment 506B corresponds to the second lane group 406B, the third road segment 506C corresponds to the third lane group 406C, and the fourth road segment 506D corresponds to the fourth lane group 406D. Additionally, the first road junction 504A on map 502 corresponds to the first junction point 402A, the second road junction 504B on map 502 corresponds to the second junction point 402B, and the third road junction 504C on map 502 corresponds to the third junction point 402C.

[0058] Return to reference Figure 1 For example, the process 100 may include: association component 114 using lane group data 128 and map data 130 to determine one or more lanes associating with roads 136. In some examples, association component 114 may attempt to match each perceived junction with the corresponding map road junction, and to match each perceived lane group with the corresponding road segment.

[0059] In some instances, to match perceived lane merging points with map road merging points, the association component 114 can simply associate the closest merging points. However, in some examples, the association component can perform a more detailed and / or robust matching algorithm / process. For example, the association component 114 can use latitude and longitude points in map data 130 to calculate turning angles between different road segments. In some examples, for road merging points with multiple successor roads (e.g., a road merging point connecting three or more road segments), the association component can calculate the turning angle between each successor road segment. The association component 114 can then use the turning angles from map data 130 to verify whether the turning angles from map data 30 satisfy a predefined pattern. If the turning angles from the perceived merging point do not conform to this pattern, the match can be rejected. Otherwise, the association component 114 can continue the matching process, and for each potential match, the association component 114 can calculate a cost score based on the merging location and the lane / road topology at the merging point, which represents the degree of matching between the perceived merging point and the map merging point. For example, the cost score can be calculated based at least on the difference between the distance from the perceived merging point to the ego machine and the distance from the map road merging point to the ego machine, and / or at least on the sum of the differences between the turning angles of the perceived merging point / lane and the map merging point / road. In some examples, the association component 114 can then select the merging point with the lowest cost score as the matching merging point.

[0060] Additionally, in some examples, the association component 114 may match or attempt to match each lane group within a lane group with its corresponding map segment. In some instances, if a matching merging point exists, the association component 114 may match a lane group with all road segments associated with that matching merging point. If no matching merging point exists, the association component 114 may match a lane group with a map segment, or avoid matching a lane group with a map segment. In some examples, the association component 114 may match lane groups with map segments based on the angle between lane groups, the distance between lane groups, the map turning angle of the road segment, the mapped form of way or classification of the road segment, or other factors. Additionally, in some instances, the association component 114 may remove lane groups with low confidence.

[0061] For example, Figure 6 The illustration shows the use of a meeting point 402 for matching according to some embodiments of the present disclosure. Figure 4 The example of the sensing lane group 406 and Figure 5Example visualization of map segment 506 in the example. As shown, the topology / location of merging point 402 can approximately correspond to the topology or location of road junction 504. Additionally, the topology of lane group 406 and the angle between lane groups 406 can approximately correspond to the topology of segment 506 and the turning angle between segments 506. Association component 114 can use these similarities between merging point 402 and road junction 504, and between lane group 406 and segment 506, to match merging point 402 with road junction 504 and lane group 406 with segment 506. For example, association component 114 can match first merging point 402A with first road junction 504A, match second merging point 402B with second road junction 504B, and match third merging point 402C with third road junction 504C. Additionally, the association component 114 can match the first lane group 406A with the first road segment 506A, the second lane group 406B with the second road segment 506B, the third lane group 406C with the third road segment 506C, and the fourth lane group 406D with the fourth road segment 506D. In some examples, one or more lanes and roads can be associated with each other. The road association 136 can indicate these matches / associations.

[0062] Return to reference Figure 1 For example, the process 100 may further include: one or more driving stack components 116 obtaining one or more lane-to-road associations 136. One or more driving stack components 116 may use the one or more lane-to-road associations 136 to perform various machine-associated operations. In some instances, the range of machine-associated operations can be from using matched perceived lanes and map roads as input to other systems or components of the machine to adjusting the machine's speed, steering angle, behavior, etc. For example, a path prediction component of one or more driving stack components 116 may use the one or more lane-to-road associations 136 to predict the machine's path, such as whether the machine (or its occupant) intends to deviate from one road segment to another. As another example, a planning component of one or more driving stack components 116 may use the one or more lane-to-road associations 136 to plan the path the machine will follow through the environment, which may indicate the specific lanes the machine will use, rather than simply indicating which road segments to use. As another example, the curvature component of one or more driving stack components can use one or more lanes and road associations 136 to calculate the curvature associated with the machine's path, and use that curvature to set operating thresholds for the machine, such as the maximum speed at which the machine can operate along different parts of the path.

[0063] Now for reference Figure 7 , Figure 7This is an example data flow diagram illustrating a process 700 for training one or more machine learning models to associate a perceived path with a corresponding map path, according to some embodiments of the present disclosure. For example, one or more machine learning models 702 may correspond to or be used to perform the functions of one or more of the perception component 102, preprocessing component 104, lane classification component 106, deviation component 108, merging component 110, grouping component 112, association component 114 and / or one or more driving stack components 116.

[0064] As shown, one or more machine learning models 702 can be trained using various input data 704 (e.g., training input data). In some examples, input data 704 may include one or more actual (e.g., previously generated and / or stored) versions of sensor data 118, perception data 120, perceived lane data 122, lane departure data 124, lane merging data 126, lane group data 128, map data 130, road merging data 132, road segment data 134, etc. Additionally or alternatively, input data 704 may be based on actual versions of sensor data 118, perception data 120, perceived lane data 122, lane departure data 124, lane merging data 126, lane group data 128, map data 130, road merging data 132, and / or road segment data 134. For example, input data 704 may include one or more modified versions of sensor data 118, perception data 120, perception lane data 122, lane departure data 124, lane merging data 126, lane group data 128, map data 130, road merging data 132 and / or road segment data 134.

[0065] One or more machine learning models 702 can be trained using input data 704 and corresponding ground truth data 706. Ground truth data 706 may include annotations, labels, masks, etc. For example, in some embodiments, ground truth data 706 may indicate the actual values ​​of parameters associated with lane merging points, lane groups, distances between lanes, and / or input data 704. For example, parameters in ground truth data 706 may include, but are not limited to, the location of lane merging points, the location of road merging points, the distance between adjacent lanes, the angle between lane groups, the angle between road segments, and / or any other parameters. In some examples, ground truth data 706 may be generated within a drawing program (e.g., annotating program), a computer-aided design (CAD) program, a labeling program, another type of program suitable for generating ground truth data 706, and / or may be hand-drawn. In any example, the ground truth data 706 can be generated synthetically (e.g., generated from a computer model or rendering), realistically (e.g., designed and generated from real-world data), machine-automated (e.g., using feature analysis and learning to extract features from the data and then generating labels), human-annotated (e.g., annotators or annotation experts define the location of labels), and / or a combination thereof (e.g., human identification of polyline vertices, machine generation of polygons using a polygon rasterizer).

[0066] Training engine 708 may use one or more loss functions that measure the loss (e.g., error) in output data 710 generated by one or more machine learning models 702 compared to ground truth data 706. Output data 710 may include lane departure data 124, lane merging data 126, lane group data 128, one or more lane-to-road associations 136, and / or any other output. In some examples, any type of loss function may be used, such as cross-entropy loss, mean squared error, mean absolute error, mean bias error, and / or other loss function types. In some examples, different outputs may have different loss functions. For example, the first perceived location of a lane merging point may include a first loss, the second perceived location of a lane merging point may include a second loss, the third perceived location of a lane merging point may include a third loss, and so on. In these examples, loss functions may be combined to form a total loss, and in some instances, training engine 708 may train one or more machine learning models 702 using the total loss by updating one or more parameters 712 (e.g., weights, biases, etc.) of one or more machine learning models 702. In any example, backpropagation computation can be performed to recursively compute the gradients of one or more loss functions with respect to the training parameters. In some examples, the weights and biases of one or more machine learning models 702 can be used to compute these gradients.

[0067] One or more machine learning models 702 may use any type of machine learning techniques and / or algorithms. For example, but not limited to, any machine learning model among the various machine learning models described herein may include any type of machine learning model, such as one or more machine learning models using linear regression, logistic regression, decision trees, support vector machines (SVM), Naive Bayes, k-nearest neighbors (Knn), k-means clustering, random forests, dimensionality reduction algorithms, gradient boosting algorithms, neural networks (e.g., autoencoders, convolutions, recurrent, perceptrons, long / short-term memory (LSTM), Hopfield, Boltzmann, deep belief, deconvolution, generative adversarial, liquid machines, large language models, visual language models, multimodal language models, diffusion, transducers, encoder-only, decoder-only, encoder-decoder, etc.) and / or other types of machine learning models.

[0068] In some examples, the machine learning model 702 can be packaged as a microservice—such as an inference microservice (e.g., NVIDIA NIM)—which may include a container (e.g., an operating system (OS)-level virtualization package) that may include an application programming interface (API) layer, a server layer, a runtime layer, and / or a model “engine.” For example, the inference microservice may include the container itself and the model 702 (e.g., weights and biases). In some instances (such as when the machine learning model 702 is small enough (e.g., has a sufficiently small number of parameters)), the model 702 may be included within the container itself. In some embodiments, the machine learning model 702 described herein can be deployed as an inference microservice to accelerate model deployment on any cloud, data center, or edge computing system while ensuring data security. For example, an inference microservice may include one or more APIs, pre-configured containers for simplified deployment, an optimized inference engine (e.g., execution software built using standardized AI model deployments, such as NVIDIA's Triton inference server), and / or one or more APIs for high-performance deep learning inference, which may include inference runtime and model optimizations for low latency and high throughput for production applications (such as NVIDIA's TensorRT) and / or enterprise management data for telemetry (e.g., including identity, metrics, health checks, and / or monitoring). One or more machine learning models 702 described herein may be included as part of a microservice along with acceleration infrastructure capable of deployment using a single command and / or orchestrated and automatically scaled on the acceleration infrastructure (e.g., on a single device up to data center scale) using a container orchestration system. Thus, an inference microservice may include one or more machine learning models 702 (e.g., already optimized for high-performance inference), inference runtime software that executes one or more machine learning models 702 and provides output / response to inputs (e.g., user queries, prompts, etc.), and enterprise management software that provides health checks, identity, and other monitoring. In some embodiments, the inference microservice may include software that performs in-situ replacement and / or updates to one or more machine learning models 702. When a replacement or update is performed, the software performing the replacement / update may maintain user configurations for the inference runtime software and the enterprise management software.

[0069] Now for reference Figure 8 , Figure 8An example of a system 802, which can perform one or more of the processes described herein, is illustrated according to some embodiments of the present disclosure. As shown, the system 802 (which may represent and / or include one or more example computing devices 1200 and / or example data centers 1300) may include one or more processors 804 (which may be similar to and / or include CPU 1206 and / or GPU 1208) and memory 806 (which may be similar to and / or include memory 1204). For example, memory 806 may store one or more of the following: perception component 102, preprocessing component 104, lane classification component 106, deviation component 108, merging component 110, grouping component 112, association component 114, one or more machine learning models 702 and / or training engines 708. Additionally, one or more processors 804 may execute one or more of the following: perception component 102, preprocessing component 104, lane classification component 106, deviation component 108, merging component 110, grouping component 112, association component 114, one or more machine learning models 702 and / or training engine 708, to perform one or more processes described herein.

[0070] For example, the system 802 may receive input data 808 generated by one or more components 810 of one or more machines 812 (e.g., one or more sensors), which may correspond to machine 1100 described herein. Input data 808 may include one or more of the following: sensor data 118, perception data 120, map data 130, and / or any other data described herein. The system 802 may then process and evaluate the input data 808 to match perceived lanes and / or lane merging points in the perception data with map segments and / or road merging points in the map data. The system 802 may send output data 814, which may include one or more lane-to-road associations 136, lane group data 128, lane merging data 126, and / or any other output data described herein. One or more drive stack components 116 of machine 812 may use the output data 814 to control one or more operations of one or more machines 812. Although depicted as a separate system, in some examples, system 802 and one or more machines 812 may be the same or different systems. For example, one or more processors 804 and memory 806 may be part of one or more machines 812 (e.g., included within a computing device of one or more machines 812).

[0071] Now for reference Figure 9 and Figure 10 Each block of methods 900 and 1000 described herein includes a computational process that can be performed using any combination of hardware, firmware, and / or software. For example, various functions can be performed by a processor executing instructions stored in memory. These methods can also be embodied as computer-usable instructions stored on a computer storage medium. These methods can be provided by standalone applications, services, or managed services (independently or in combination with other managed services) or plug-ins of another product, to name a few. Additionally, as an example regarding... Figure 1 Methods 900 and 1000 are described. However, additionally or alternatively, these methods may be performed by any one of the systems or any combination of systems, including but not limited to the systems described herein.

[0072] Figure 9 This is a flowchart illustrating an example of a method 900 for associating perceived lanes with map driveways according to some embodiments of the present disclosure. At block B902, the method 900 may include: determining one or more first locations in the environment corresponding to one or more first merging points, at which one or more first lanes deviate from one or more second lanes. For example, merging component 110 may determine one or more first locations in the environment corresponding to one or more first merging points, at which one or more first lanes deviate from one or more second lanes.

[0073] At box B904, method 900 may include: determining one or more second locations in the environment corresponding to one or more second junction points, at which one or more first road segments deviate from one or more second road segments. For example, association component 114 may determine one or more second locations in the environment corresponding to one or more second junction points, at which one or more first road segments deviate from one or more second road segments. In some examples, association component 114 may use map data to determine one or more second locations. For example, map data may indicate one or more second locations corresponding to one or more second junction points.

[0074] At box B906, method 900 may include: using at least one or more first locations and one or more second locations to associate one or more first lanes with one or more first road segments and to associate one or more second lanes with one or more second road segments. For example, association component 114 may associate one or more first lanes with one or more first road segments and to associate one or more second lanes with one or more second road segments. Association component 114 may use one or more first locations and one or more second locations to associate lanes with road segments. For example, association component 114 may determine which of the one or more first locations and one or more second locations are closest to each other and use the proximity between these locations to match perceived merging points and lane-to-map road merging points and roads.

[0075] At box B908, the method 900 may include performing one or more operations associated with a machine in the environment, based at least on the association. For example, one or more driving stack components 116 may perform one or more operations associated with a machine in the environment, based at least on one or more lane and road associations 136. In some examples, one or more operations may include: changing the machine's trajectory, adjusting the machine's speed, setting one or more operational constraints for the machine, planning a path for the machine to follow, etc.

[0076] Now for reference Figure 10 , Figure 10 This is a flowchart illustrating an example of a method 1000 for associating a sensed path with a map path using merging points, according to some embodiments of the present disclosure. At block B1002, the method 1000 may include: determining one or more locations in the environment corresponding to one or more merging points associated with one or more first sensed paths and one or more second sensed paths. For example, merging component 110 may determine one or more locations in the environment corresponding to one or more merging points associated with one or more first sensed paths and one or more second sensed paths. In some examples, the sensed path may correspond to a sensed lane in the environment.

[0077] At box B1004, method 1000 may include: evaluating one or more locations based at least on map data representing an environment map to determine that one or more first-perceived paths correspond to at least one map path. For example, association component 114 may determine that one or more first-perceived paths correspond to at least one map path. In some examples, the map path may correspond to a map road segment in the environment. In some examples, one or more locations corresponding to one or more junctions may be compared with one or more second locations corresponding to map road junctions associated with the map path to associate one or more first-perceived paths with the map path.

[0078] At box B1006, the method 1000 may include performing one or more operations associated with a machine in the environment, based at least on associating one or more first perceived paths with at least one map path. For example, one or more driving stack components 116 may perform one or more operations associated with a machine in the environment, based at least on one or more lane and road associations 136. In some examples, one or more operations may include: predicting the path the machine intends to follow, calculating the curvature associated with the predicted path, planning the path or trajectory the machine will follow (e.g., planning a more detailed path based on the association of one or more perceived paths with map paths), adjusting the machine's operating parameters (e.g., setting the maximum turning speed based on the calculated curvature), etc.

[0079] Example autonomous vehicles

[0080] Figure 11AThis is an illustration of an example autonomous vehicle 1100 according to some embodiments of the present disclosure. The autonomous vehicle 1100 (or, alternatively, referred to herein as “vehicle 1100”) may include, but is not limited to, passenger vehicles such as cars, trucks, buses, first-response vehicles, shuttles, electric or motorized bicycles, motorcycles, fire trucks, police vehicles, ambulances, boats, engineering vehicles, submarines, robotic vehicles, drones, aircraft, vehicles coupled to trailers (e.g., semi-trailer trucks for hauling goods) and / or other types of vehicles (e.g., driverless and / or vehicles accommodating one or more passengers). Autonomous vehicles are typically described according to the levels of automation defined by the National Highway Traffic Safety Administration (NHTSA), a division of the U.S. Department of Transportation, and the Society of Automotive Engineers (SAE) in its "Taxonomy and Definitions for Terms Related to Driving Automation Systems for On-Road Motor Vehicles" (Standard No. J3016-201806, published June 15, 2018; Standard No. J3016-201609, published September 30, 2016; and previous and future versions of this standard). Vehicle 1100 may be able to implement one or more functions that meet Level 3-5 of the autonomous driving level. Vehicle 1100 may be able to implement one or more functions that meet Level 1-5 of the autonomous driving level. For example, depending on the embodiment, vehicle 1100 may be able to implement driver assistance (Level 1), partial automation (Level 2), conditional automation (Level 3), high automation (Level 4), and / or full automation (Level 5). As used herein, the term “autonomy” can include any and / or all types of autonomy of the vehicle 1100 or other machine, such as full autonomy, high autonomy, conditional autonomy, partial autonomy, assisted autonomy, semi-autonomy, primary autonomy or other designations.

[0081] Vehicle 1100 may include components such as chassis, body, wheels (e.g., 2, 4, 6, 8, 18, etc.), tires, axles, and other vehicle components. Vehicle 1100 may include a propulsion system 1150, such as an internal combustion engine, a hybrid power plant, an all-electric motor, and / or another type of propulsion system. Propulsion system 1150 may be connected to the drivetrain of vehicle 1100, which may include a transmission, to enable propulsion of vehicle 1100. Propulsion system 1150 may be controlled in response to receiving a signal from throttle / accelerator 1152.

[0082] A steering system 1154, which may include a steering wheel, can be used to steer the vehicle 1100 (e.g., along a desired path or route) when the propulsion system 1150 is operating (e.g., when the vehicle is in motion). The steering system 1154 may receive signals from the steering actuator 1156. For fully automatic (level 5) functionality, the steering wheel may be optional.

[0083] The brake sensor system 1146 can be used to operate the vehicle brakes in response to receiving signals from the brake actuator 1148 and / or the brake sensor.

[0084] It may include one or more System-on-Chip (SoC) 1104 ( Figure 11C One or more controllers 1136, including and / or one or more GPUs, may provide signals (e.g., signals representing commands) to one or more components and / or systems of vehicle 1100. For example, one or more controllers may send signals to operate vehicle brakes via one or more brake actuators 1148, to operate steering system 1154 via one or more steering actuators 1156, and to operate propulsion system 1150 via one or more throttles / accelerators 1152. One or more controllers 1136 may include one or more onboard (e.g., integrated) computing devices (e.g., supercomputers) that process sensor signals and output operating commands (e.g., signals representing commands) to enable autonomous driving and / or assist a human driver in driving vehicle 1100. One or more controllers 1136 may include a first controller 1136 for autonomous driving functions, a second controller 1136 for functional safety functions, a third controller 1136 for artificial intelligence functions (e.g., computer vision), a fourth controller 1136 for infotainment functions, a fifth controller 1136 for redundancy in emergency situations, and / or other controllers. In some examples, a single controller 1136 can handle two or more of the functions described above, and two or more controllers 1136 can handle a single function, and / or any combination thereof.

[0085] One or more controllers 1136 may provide signals for controlling one or more components and / or systems of vehicle 1100 in response to sensor data (e.g., sensor inputs) received from one or more sensors. Sensor data may be received from, for example, but not limited to, Global Navigation Satellite System (“GNSS”) sensors 1158 (e.g., Global Positioning System sensors), RADAR sensors 1160, ultrasonic sensors 1162, LIDAR sensors 1164, inertial measurement unit (IMU) sensors 1166 (e.g., accelerometers, gyroscopes, magnetic compasses, magnetometers, etc.), microphones 1196, stereo cameras 1168, wide-angle cameras 1170 (e.g., fisheye cameras), infrared cameras 1172, surround cameras 1174 (e.g., 360-degree cameras), long-range and / or medium-range cameras 1198, speed sensors 1144 (e.g., for measuring the rate of vehicle 1100), vibration sensors 1142, steering sensors 1140, braking sensors (e.g., as part of braking sensor system 1146), and / or other sensor types.

[0086] One or more of the controllers 1136 may receive inputs (e.g., represented by input data) from the instrument cluster 1132 of the vehicle 1100 and provide outputs (e.g., represented by output data, display data, etc.) via a human-machine interface (HMI) display 1134, an auditory signaling device, a speaker, and / or via other components of the vehicle 1100. These outputs may include information such as vehicle speed, rate, time, map data (e.g., [missing information]). Figure 11C Information such as a high-definition (“HD”) map 1122, location data (e.g., the location of vehicle 1100 on the map), direction, the location of other vehicles (e.g., occupying a grid), and information about objects and their states perceived by controller 1136, etc. For example, HMI display 1134 may display information about the existence of one or more objects (e.g., street signs, warning signs, traffic light changes, etc.) and / or information about driving maneuvers that the vehicle has made, is making, or will make (e.g., changing lanes now, leaving 34B in two miles, etc.).

[0087] Vehicle 1100 also includes a network interface 1124, which can communicate via one or more networks using one or more wireless antennas 1126 and / or a modem. For example, network interface 1124 may be able to communicate via Long Term Evolution (“LTE”), Wideband Code Division Multiple Access (“WCDMA”), Universal Mobile Telecommunications System (“UMTS”), Global System for Mobile Communications (“GSM”), IMT-CDMA Multicarrier (“CDMA2000”), etc. One or more wireless antennas 1126 may also enable communication between objects in the environment (e.g., vehicles, mobile devices, etc.) using one or more local area networks such as Bluetooth, Bluetooth Low Energy (“LE”), Z-Wave, ZigBee, etc., and / or one or more low-power wide area networks (LPWANs such as LoRaWAN, SigFox, etc.).

[0088] Figure 11B For use in accordance with some embodiments of this disclosure Figure 11A This is an example of the camera position and field of view of an example autonomous vehicle 1100. The camera and its respective field of view are an example embodiment and are not intended to be limiting. For example, additional and / or replaceable cameras may be included, and / or these cameras may be located at different positions on the vehicle 1100.

[0089] The camera type used for the camera may include, but is not limited to, a digital camera suitable for use with components and / or systems of vehicle 1100. The camera may operate at Automotive Safety Integrity Level (ASIL) B and / or another ASIL. The camera type may have any image capture rate, such as 60 frames per second (fps), 120 fps, 240 fps, etc., depending on the embodiment. The camera may be able to use a rolling shutter, a global shutter, another type of shutter, or a combination thereof. In some examples, the color filter array may include a red-white-white-white (RCCC) color filter array, a red-white-white-blue (RCCB) color filter array, a red-blue-green-white (RBGC) color filter array, a Foveon X3 color filter array, a Bayer sensor (RGGB) color filter array, a monochrome sensor color filter array, and / or another type of color filter array. In some embodiments, a sharp-pixel camera, such as a camera with RCCC, RCCB, and / or RBGC color filter arrays, may be used in efforts to improve light sensitivity.

[0090] In some examples, one or more of the cameras can be used to perform advanced driver assistance system (ADAS) functions (e.g., as part of a redundant or fail-safe design). For example, a multi-function monocular camera can be installed to provide functions including lane departure warning, traffic sign assistance, and intelligent headlight control. One or more of the cameras (e.g., all cameras) can simultaneously record and provide image data (e.g., video).

[0091] One or more of the cameras can be mounted in mounting components such as custom-designed (3D-printed) components to cut off stray light and reflections from inside the vehicle (e.g., reflections from the dashboard reflected in the windshield mirror) that may interfere with the camera's image data capture capabilities. Regarding the wing mirror mounting components, the wing mirror components can be custom-3D printed so that the camera mounting plate matches the shape of the wing mirror. In some examples, one or more cameras can be integrated into the wing mirror. For side-view cameras, one or more cameras can also be integrated into the four pillars at each corner of the cab.

[0092] A camera with a field of view that includes the environment in front of the vehicle 1100 (e.g., a front-facing camera) can be used for surround view to help identify forward paths and obstacles, and, with the assistance of one or more controllers 1136 and / or control SoCs, to provide information crucial for generating an occupancy grid and / or determining a preferred vehicle path. The front-facing camera can be used to perform many of the same ADAS functions as LiDAR, including emergency braking, pedestrian detection, and collision avoidance. The front-facing camera can also be used in ADAS functions and systems, including Lane Departure Warning (“LDW”), Autonomous Cruise Control (“ACC”), and / or other functions such as traffic sign recognition.

[0093] A variety of cameras can be used in front-facing configurations, including, for example, monocular camera platforms including complementary metal-oxide-semiconductor (“CMOS”) color imagers. Another example could be a wide-angle camera 1170, which can be used to perceive objects entering the field of view from the periphery (such as pedestrians, traffic at intersections, or bicycles). Although Figure 11B The diagram shows only one wide-angle camera, but any number (including zero) of wide-angle cameras 1170 can be present on vehicle 1100. Furthermore, any number of one or more remote cameras 1198 (e.g., a pair of long-view stereo cameras) can be used for depth-based object detection, especially for objects for which a neural network has not yet been trained. One or more remote cameras 1198 can also be used for object detection and classification, as well as basic object tracking.

[0094] Any number of stereo cameras 1168 may also be included in a front-mounted configuration. In at least one embodiment, one or more stereo cameras 1168 may include an integrated control unit comprising a scalable processing unit that can provide a multi-core microprocessor and programmable logic (“FPGA”) with an integrated controller area network (“CAN”) or Ethernet interface on a single chip. Such a unit can be used to generate a 3D map of the vehicle environment, including distance estimates for all points in the image. Alternative stereo cameras 1168 may include a compact stereo vision sensor that may include two camera lenses (one on each side) and an image processing chip capable of measuring the distance from the vehicle to a target object and using the generated information (e.g., metadata) to activate autonomous emergency braking and lane departure warning functions. Other types of stereo cameras 1168 may be used in addition to those described herein or alternatively.

[0095] A camera (e.g., a side-view camera) having a field of view that includes the side of the vehicle 1100 can be used for surround view, providing information for creating and updating occupancy grids and generating side-impact collision warnings. For example, a surround camera 1174 (e.g., ...) Figure 11B The four surround cameras 1174 shown can be mounted on vehicle 1100. The surround cameras 1174 can include wide-angle cameras 1170, fisheye cameras, 360-degree cameras, and / or the like. Four examples are provided; the four fisheye cameras can be positioned at the front, rear, and sides of the vehicle. In an alternative arrangement, the vehicle can use three surround cameras 1174 (e.g., left, right, and rear) and can utilize one or more other cameras (e.g., forward-facing cameras) as a fourth surround-view camera.

[0096] A camera (e.g., a rear-view camera) having a field of view that includes the environment behind the vehicle 1100 can be used for parking assistance, surround view, rear collision warning, and creating and updating occupancy grids. A wide variety of cameras can be used, including but not limited to those also suitable as front-facing cameras as described herein (e.g., long-range and / or mid-range camera 1198, stereo camera 1168, infrared camera 1172, etc.).

[0097] Figure 11C For use in accordance with some embodiments of this disclosure Figure 11AThe example autonomous vehicle 1100 is illustrated in the block diagram of an example system architecture. It should be understood that this arrangement, and other arrangements described herein, are merely illustrative. Other arrangements and elements (e.g., machines, interfaces, functions, sequences, functional groupings, etc.) may be used in addition to or in place of those shown, and some elements may be omitted entirely. Furthermore, many of the elements described herein are functional entities, which may be implemented as discrete or distributed components or in combination with other components, and in any suitable combination and location. The various functions described herein as being performed by these entities can be implemented via hardware, firmware, and / or software. For example, the various functions can be implemented by a processor executing instructions stored in memory.

[0098] Figure 11C Each component, feature, and system in vehicle 1100 is illustrated as being connected via bus 1102. Bus 1102 may include a Controller Area Network (CAN) data interface (or, alternatively, referred to herein as the "CAN bus"). CAN may be a network within vehicle 1100 used to assist in the control of various features and functions of vehicle 1100, such as the actuation of brakes, acceleration, braking, steering, windshield wipers, etc. The CAN bus may be configured to have dozens or even hundreds of nodes, each with its own unique identifier (e.g., CAN ID). The CAN bus can be read to find steering wheel angle, ground speed, engine speed per minute (RPM), button positions, and / or other vehicle status indicators. The CAN bus may be ASIL B compliant.

[0099] Although bus 1102 is described herein as a CAN bus, this is not intended to be limiting. For example, FlexRay and / or Ethernet may be used in addition to or alternatively to a CAN bus. Furthermore, although bus 1102 is represented by a single line, this is not intended to be limiting. For example, any number of buses 1102 may exist, which may include one or more CAN buses, one or more FlexRay buses, one or more Ethernet buses, and / or one or more other types of buses using different protocols. In some examples, two or more buses 1102 may be used to perform different functions and / or may be used for redundancy. For example, a first bus 1102 may be used for a collision avoidance function, and a second bus 1102 may be used for drive control. In any example, each bus 1102 may communicate with any component of vehicle 1100, and two or more buses 1102 may communicate with the same component. In some examples, each SoC 1104, each controller 1136, and / or each computer within the vehicle may have access to the same input data (e.g., input from sensors in the vehicle 1100) and may be connected to a common bus such as the CAN bus.

[0100] Vehicle 1100 may include one or more controllers 1136, such as those described herein. Figure 11A The controllers described herein. Controller 1136 can be used for a wide variety of functions. Controller 1136 can be coupled to any other different components and systems of vehicle 1100 and can be used for the control of vehicle 1100, artificial intelligence of vehicle 1100, infotainment and / or the like for vehicle 1100.

[0101] Vehicle 1100 may include one or more System-on-Chip (SoC) 1104. SoC 1104 may include CPU 1106, GPU 1108, processor 1110, cache 1112, accelerator 1114, data storage 1116, and / or other components and features not shown. SoC 1104 can be used to control vehicle 1100 across a wide variety of platforms and systems. For example, one or more SoCs 1104 may be combined with an HD map 1122 in a system (e.g., the system of vehicle 1100), the HD map being accessible from one or more servers (e.g., via a network interface 1124). Figure 11D One or more servers (1178) receive map refresh and / or updates.

[0102] CPU 1106 may include CPU clusters or CPU complexes (or, alternatively, referred to herein as "CCPLEX"). CPU 1106 may include multiple cores and / or L2 cache. For example, in some embodiments, CPU 1106 may include eight cores in a coherent multiprocessor configuration. In some embodiments, CPU 1106 may include four dual-core clusters, each cluster having a dedicated L2 cache (e.g., 2MB L2 cache). CPU 1106 (e.g., CCPLEX) may be configured to support simultaneous cluster operation, such that any combination of clusters of CPU 1106 can be active at any given time.

[0103] CPU 1106 can implement power management capabilities including one or more of the following features: automatic clock gating of hardware blocks when idle to save dynamic power; clock gating of each core when the core is not actively executing instructions due to the execution of WFI / WFE instructions; independent power gating of each core; independent clock gating of each core cluster when all cores are clock-gated or power-gated; and / or independent power gating of each core cluster when all cores are power-gated. CPU 1106 can further implement enhanced algorithms for managing power states, wherein allowed power states and desired wake-up times are specified, and the hardware / microcode determines the optimal power state to enter for the core, cluster, and CCPLEX. The processing core can support simplified power state entry sequences in software, with this work offloaded to the microcode.

[0104] GPU 1108 may include an integrated GPU (or, alternatively, referred to herein as an "iGPU"). GPU 1108 may be programmable and efficient for parallel workloads. In some examples, GPU 1108 may use an enhanced tensor instruction set. GPU 1108 may include one or more streaming microprocessors, wherein each streaming microprocessor may include an L1 cache (e.g., an L1 cache with at least 96KB of storage capacity), and two or more of these streaming microprocessors may share an L2 cache (e.g., an L2 cache with 512KB of storage capacity). In some embodiments, GPU 1108 may include at least eight streaming microprocessors. GPU 1108 may use a computation application programming interface (API). Furthermore, GPU 1108 may use one or more parallel computing platforms and / or programming models (e.g., NVIDIA's CUDA).

[0105] In automotive and embedded applications, the GPU 1108 can be power-optimized for optimal performance. For example, the GPU 1108 can be fabricated on FinFETs. However, this is not intended to be limiting, and the GPU 1108 can be fabricated using other semiconductor manufacturing processes. Each streaming microprocessor can combine several mixed-precision processing cores divided into multiple blocks. For example, and without limitation, 64 PF32 cores and 32 PF64 cores can be divided into four processing blocks. In such an example, each processing block can be allocated 16 FP32 cores, 8 FP64 cores, 16 INT32 cores, two mixed-precision NVIDIA Tensor cores for deep learning matrix arithmetic, an L0 instruction cache, a warp scheduler, a dispatch unit, and / or a 64KB register file. Furthermore, the streaming microprocessor can include independent parallel integer and floating-point data paths to leverage the mixture of computation and addressing computations to provide efficient execution of workloads. Streaming microprocessors may include independent thread scheduling capabilities to allow for finer-grained synchronization and cooperation between parallel threads. Streaming microprocessors may include combined L1 data caches and shared memory units to improve performance while simplifying programming.

[0106] The GPU 1108 may include, in some examples, a High Bandwidth Memory (HBM) and / or a 16GB HBM2 memory subsystem providing a peak memory bandwidth of approximately 900GB / s. In some examples, in addition to HBM memory or alternatively, Synchronous Graphics Random Access Memory (SGRAM), such as Generation 5 Graphics Double Data Rate Synchronous Random Access Memory (GDDR5), may be used.

[0107] The GPU 1108 may include unified memory technology, which includes access counters to allow memory pages to be migrated more precisely to the processors that access them most frequently, thereby improving the efficiency of shared memory ranges between processors. In some examples, Address Translation Service (ATS) support can be used to allow the GPU 1108 to directly access the CPU 1106 page tables. In such examples, when the GPU 1108 Memory Management Unit (MMU) experiences a miss, the address translation request can be transferred to the CPU 1106. In response, the CPU 1106 can look up the virtual-physical mapping for the address in its page tables and transfer the translation back to the GPU 1108. Thus, unified memory technology can allow a single unified virtual address space for the memory of both the CPU 1106 and the GPU 1108, simplifying GPU 1108 programming and porting applications to the GPU 1108.

[0108] In addition, the GPU 1108 may include access counters that track how frequently the GPU 1108 accesses the memory of other processors. Access counters can help ensure that memory pages are moved to the physical memory of the processor that accesses those pages most frequently.

[0109] SoC 1104 may include any number of caches 1112, including those described herein. For example, cache 1112 may include an L3 cache available to both CPU 1106 and GPU 1108 (e.g., it is connected to both CPU 1106 and GPU 1108). Cache 1112 may include a write-back cache, which can track the state of rows, for example, using a cache coherence protocol (e.g., MEI, MESI, MSI, etc.). Depending on the embodiment, the L3 cache may include 4 MB or more, but a smaller cache size may also be used.

[0110] SoC 1104 may include an arithmetic logic unit (ALU) that can be utilized in processing of any of the various tasks or operations performed on vehicle 1100, such as processing a DNN. Furthermore, SoC 1104 may include a floating-point unit (FPU) (or other mathematical coprocessor or digital coprocessor type) for performing mathematical operations within the system. For example, SoC 1104 may include one or more FPUs integrated as execution units within CPU 1106 and / or GPU 1108.

[0111] SoC 1104 may include one or more accelerators 1114 (e.g., hardware accelerators, software accelerators, or combinations thereof). For example, SoC 1104 may include a hardware accelerator cluster, which may include optimized hardware accelerators and / or large on-chip memory. This large on-chip memory (e.g., 4MB SRAM) can enable the hardware accelerator cluster to accelerate neural networks and other computations. The hardware accelerator cluster can be used to complement GPU 1108 and offload some tasks from GPU 1108 (e.g., freeing up more cycles of GPU 1108 to perform other tasks). As an example, accelerator 1114 can be used for targeted workloads (e.g., perceptrons, convolutional neural networks (CNNs), etc.) that are sufficiently stable to allow for easy control of acceleration. When used herein, the term "CNN" can include all types of CNNs, including region-based or region convolutional neural networks (RCNNs) and fast RCNNs (e.g., for object detection).

[0112] Accelerator 1114 (e.g., a hardware accelerator cluster) may include a Deep Learning Accelerator (DLA). The DLA may include one or more Tensor Processing Units (TPUs) that can be configured to provide an additional 10 trillion operations per second for deep learning applications and inference. The TPU may be an accelerator configured to perform image processing functions (e.g., for CNNs, RCNNs, etc.) and optimized for performing image processing functions. The DLA may be further optimized for a specific set of neural network types and floating-point operations and inference. The DLA is designed to provide higher performance per millimeter than a general-purpose GPU and significantly outperform CPUs. The TPU can perform several functions, including single-instance convolution functions, support for INT8, INT16, and FP16 data types for both features and weights, and post-processor functions.

[0113] DLA can execute neural networks, especially CNNs, quickly and efficiently on processed or unprocessed data for any function across a wide variety of applications, such as, but not limited to: CNNs for object recognition and detection using data from camera sensors; CNNs for distance estimation using data from camera sensors; CNNs for emergency vehicle detection and recognition using data from microphones; CNNs for face recognition and vehicle owner recognition using data from camera sensors; and / or CNNs for safety and / or safety-related events.

[0114] The DLA can perform any function of the GPU 1108, and by using inference accelerators, for example, designers can make the DLA or GPU 1108 target any function. For example, designers can focus the CNN processing and floating-point operations on the DLA and leave other functions to the GPU 1108 and / or other accelerators 1114.

[0115] Accelerator 1114 (e.g., a cluster of hardware accelerators) may include a programmable vision accelerator (PVA), which may alternatively be referred to herein as a computer vision accelerator. The PVA may be designed and configured to accelerate computer vision algorithms for advanced driver assistance systems (ADAS), autonomous driving, and / or augmented reality (AR) and / or virtual reality (VR) applications. The PVA can provide a balance between performance and flexibility. For example, each PVA may include, for example, but not limited to, any number of reduced instruction set computer (RISC) cores, direct memory access (DMA), and / or any number of vector processors.

[0116] RISC cores can interact with image sensors (such as the image sensor of any camera described herein), image signal processors, and / or the like. Each of these RISC cores may include any amount of memory. Depending on the embodiment, the RISC core may use any of several protocols. In some examples, the RISC core may execute a real-time operating system (RTOS). RISC cores may be implemented using one or more integrated circuit devices, application-specific integrated circuits (ASICs), and / or memory devices. For example, a RISC core may include an instruction cache and / or tightly coupled RAM.

[0117] DMA enables PVA components to access system memory independently of the CPU 1106. DMA can support any number of features to provide optimizations to the PVA, including but not limited to support for multidimensional addressing and / or circular addressing. In some examples, DMA can support addressing in up to six or more dimensions, which can include block width, block height, block depth, horizontal block step, vertical block step, and / or depth step.

[0118] A vector processor can be a programmable processor designed to efficiently and flexibly execute programming for computer vision algorithms and provide signal processing capabilities. In some examples, a PVA may include a PVA core and two vector processing subsystem partitions. The PVA core may include a processor subsystem, one or more DMA engines (e.g., two DMA engines), and / or other peripherals. The vector processing subsystem may operate as the main processing engine of the PVA and may include a vector processing unit (VPU), an instruction cache, and / or vector memory (e.g., VMEM). The VPU core may include a digital signal processor, such as, for example, a Single Instruction Multiple Data (SIMD) or Very Long Instruction Word (VLIW) digital signal processor. The combination of SIMD and VLIW can enhance throughput and speed.

[0119] Each of the vector processors may include an instruction cache and may be coupled to dedicated memory. As a result, in some examples, each of the vector processors may be configured to execute independently of other vector processors. In other examples, the vector processors included in a particular PVA may be configured to employ data parallelism. For example, in some embodiments, multiple vector processors included in a single PVA may execute the same computer vision algorithm, but on different regions of an image. In other examples, vector processors included in a particular PVA may execute different computer vision algorithms simultaneously on the same image, or even different algorithms on a sequence of images or portions of an image. Among other things, any number of PVAs may be included in a cluster of hardware accelerators, and any number of vector processors may be included in each of these PVAs. Furthermore, the PVA may include additional error correction code (ECC) memory to enhance overall system security.

[0120] Accelerator 1114 (e.g., a hardware accelerator cluster) may include an on-chip computer vision network and SRAM to provide high-bandwidth, low-latency SRAM for accelerator 1114. In some examples, on-chip memory may include at least 4MB of SRAM consisting of, for example, but not limited to, eight field-configurable memory blocks, accessible by both the PVA and DLA. Each pair of memory blocks may include an Advanced Peripheral Bus (APB) interface, configuration circuitry, a controller, and a multiplexer. Any type of memory may be used. The PVA and DLA may access memory via a backbone that provides high-speed memory access to the PVA and DLA. The backbone may include (e.g., using an APB) an on-chip computer vision network that interconnects the PVA and DLA to memory.

[0121] On-chip computer vision networks can include interfaces that ensure both the PVA and DLA provide ready and valid signals before transmitting any control signals / addresses / data. Such interfaces can provide separate phases and channels for transmitting control signals / addresses / data, as well as burst communication for continuous data transmission. This type of interface can conform to ISO 26262 or IEC 61508 standards, but other standards and protocols can also be used.

[0122] In some examples, SoC 1104 may include, for example, a real-time ray tracing hardware accelerator as described in U.S. Patent Application No. 16 / 101,232, filed August 10, 2018. This real-time ray tracing hardware accelerator can be used to quickly and efficiently determine the location and extent of objects (e.g., within a world model) to generate real-time visualization simulations for RADAR signal interpretation, sound propagation synthesis and / or analysis, SONAR system simulation, general wave propagation simulation, comparison with LiDAR data for localization and / or other functional purposes, and / or other uses. In some embodiments, one or more Tree Traversal Units (TTUs) may be used to perform one or more ray tracing-related operations.

[0123] Accelerators 1114 (e.g., hardware accelerator clusters) have broad applications in autonomous driving. PVAs can be programmable vision accelerators used in critical processing stages of ADAS and autonomous vehicles. The capabilities of PVAs are a good match for algorithmic domains requiring predictable processing, low power, and low latency. In other words, PVAs perform well in semi-dense or dense rule computation, even on small datasets requiring predictable runtimes with low latency and low power. Therefore, in the context of platforms for autonomous vehicles, PVAs are designed to run classical computer vision algorithms because they are effective in object detection and integer mathematical operations.

[0124] For example, according to one embodiment of this technology, PVA is used to perform computer stereo vision. In some examples, semi-global matching-based algorithms may be used, but this is not intended to be limiting. Many applications for Level 3-5 autonomous driving require instantaneous motion estimation / stereo matching (e.g., from moving structures, pedestrian recognition, lane detection, etc.). PVA can perform computer stereo vision functions on input from two monocular cameras.

[0125] In some examples, PVA can be used to perform intensive optical flow, providing processed RADAR data from the raw RADAR data (e.g., using 4D Fast Fourier Transform). In other examples, PVA is used for time-of-flight depth processing, which, for example, involves processing raw time-of-flight data to provide processed time-of-flight data.

[0126] DLA can be used to run any type of network to enhance control and driving safety, including, for example, neural networks that output a confidence metric for each object detection. Such a confidence value can be interpreted as a probability or as providing a relative “weight” for each detection compared to other detections. This confidence value allows the system to make further decisions about which detections should be considered true positives rather than false positives. For example, the system can set a threshold for the confidence and only consider detections exceeding the threshold as true positives. In an Automatic Emergency Braking (AEB) system, false positives can cause the vehicle to automatically perform emergency braking, which is clearly undesirable. Therefore, only the most confident detections should be considered as triggers for AEB. DLA can run neural networks to regress the confidence values. The neural network can take at least some subset of parameters as its input, such as bounding box dimensions, ground plane estimates obtained (e.g. from another subsystem), outputs from inertial measurement unit (IMU) sensors 1166 related to the orientation and distance of vehicle 1100, 3D position estimates of objects obtained from the neural network and / or other sensors (e.g., LiDAR sensor 1164 or RADAR sensor 1160), etc.

[0127] SoC 1104 may include one or more data storage units 1116 (e.g., memory). Data storage units 1116 may be on-chip memory of SoC 1104, which may store neural networks to be executed on the GPU and / or DLA. In some examples, for redundancy and security, data storage units 1116 may be large enough to store multiple instances of the neural network. Data storage units 1112 may include L2 or L3 cache 1112. References to data storage units 1116 may include references to memory associated with PVA, DLA, and / or other accelerators 1114 as described herein.

[0128] SoC 1104 may include one or more processors 1110 (e.g., embedded processors). Processor 1110 may include a startup and power management processor, which may be a dedicated processor and subsystem for handling startup power and management functions, as well as safety implementation. The startup and power management processor may be part of the SoC 1104 startup sequence and may provide runtime power management services. The startup power and management processor may provide clock and voltage programming, auxiliary system low-power state transitions, SoC 1104 thermal and temperature sensor management, and / or SoC 1104 power state management. Each temperature sensor may be implemented as a ring oscillator whose output frequency is proportional to the temperature, and SoC 1104 may use the ring oscillator to detect the temperature of CPU 1106, GPU 1108, and / or accelerator 1114. If it is determined that the temperature exceeds a threshold, the startup and power management processor may enter a temperature fault routine and place SoC 1104 into a lower power state and / or place vehicle 1100 into a driver-safe parking mode (e.g., safely stop vehicle 1100).

[0129] The processor 1110 may also include a set of embedded processors that can be used as an audio processing engine. The audio processing engine can be an audio subsystem that allows for full hardware support for multi-channel audio via multiple interfaces and a wide and flexible range of audio I / O interfaces. In some examples, the audio processing engine is a dedicated processor core with a digital signal processor and dedicated RAM.

[0130] The processor 1110 may also include an always-on-processor engine that can provide the necessary hardware features to support low-power sensor management and wake-up use cases. This always-on-processor engine may include a processor core, tightly coupled RAM, support for peripherals (such as timers and interrupt controllers), various I / O controller peripherals, and routing logic.

[0131] The processor 1110 may also include a security cluster engine, which includes a dedicated processor subsystem for handling security management for automotive applications. The security cluster engine may include two or more processor cores, tightly coupled RAM, support for peripheral devices (e.g., timers, interrupt controllers, etc.), and / or routing logic. In secure mode, the two or more cores may operate in lockstep mode and function as a single core with comparison logic that detects any differences between their operations.

[0132] The processor 1110 may also include a real-time camera engine, which may include a dedicated processor subsystem for handling real-time camera management.

[0133] The processor 1110 may also include a high dynamic range signal processor, which may include an image signal processor, which is a hardware engine that is part of the camera processing pipeline.

[0134] Processor 1110 may include a video image compositer, which may be (e.g., implemented on a microprocessor) a processing block, implementing video post-processing functions required by the video playback application to generate the final image for the player window. The video image compositer may perform lens distortion correction on the wide-angle camera 1170, the surround camera 1174, and / or the in-cabin monitoring camera sensor. The in-cabin monitoring camera sensor is preferably monitored by a neural network running on another instance of an advanced SoC, configured to recognize in-cabin events and respond accordingly. The in-cabin system may perform lip reading to activate mobile phone services and make calls, dictate emails, change vehicle destinations, activate or change the vehicle's infotainment system and settings, or provide voice-activated web browsing. Some functions are only available to the driver when the vehicle is operating in autonomous mode and are disabled in other situations.

[0135] Video image compositers can include enhanced temporal denoising for both spatial and temporal noise reduction. For example, in the case of motion in the video, denoising appropriately weights spatial information, reducing the weight of information provided by neighboring frames. In cases where the image or part of the image does not contain motion, the temporal denoising performed by the video image compositer can use information from previous images to reduce noise in the current image.

[0136] The video image compositer can also be configured to perform stereo correction on input stereo lens frames. When the operating system desktop is in use and the GPU 1108 does not need to continuously render new surfaces, the video image compositer can be further used for user interface components. Even when the GPU 1108 is powered on and active, performing 3D rendering, the video image compositer can be used to offload the GPU 1108 to improve performance and responsiveness.

[0137] The SoC 1104 may also include a Mobile Industry Processor Interface (MIPI) camera serial interface, a high-speed interface, and / or a video input block that can be used for camera and related pixel input functions for receiving video and input from a camera. The SoC 1104 may also include an input / output controller that can be software controlled and can be used to receive I / O signals not assigned to a specific role.

[0138] SoC 1104 may also include a wide range of peripheral interfaces to enable communication with peripherals, audio codecs, power management and / or other devices. SoC 1104 can be used to process data from cameras and sensors (e.g., LIDAR sensor 1164, RADAR sensor 1160, etc., which can be connected via Gigabit Multimedia Serial Link and Ethernet), data from bus 1102 (e.g., vehicle 1100 speed, steering wheel position, etc.), and data from GNSS sensor 1158 (connected via Ethernet or CAN bus). SoC 1104 may also include a dedicated high-performance, high-capacity memory controller, which may include its own DMA engine and can be used to free up CPU 1106 from routine data management tasks.

[0139] The SoC 1104 can be an end-to-end platform with a flexible architecture spanning Automation Levels 3-5, providing a comprehensive functional safety architecture that leverages and efficiently utilizes computer vision and ADAS technologies for diversity and redundancy, along with deep learning tools to deliver a flexible and reliable driving software stack. The SoC 1104 can be faster, more reliable, and even more energy- and space-efficient than conventional systems. For example, when combined with the CPU 1106, GPU 1108, and data storage 1116, the accelerator 1114 can provide a fast and efficient platform for Level 3-5 autonomous vehicles.

[0140] Therefore, this technology offers capabilities and functionalities that cannot be achieved through conventional systems. For example, computer vision algorithms can be executed on CPUs, which can be configured using high-level programming languages ​​such as C to execute a wide variety of processing algorithms across a diverse range of visual data. However, CPUs often cannot meet the performance requirements of many computer vision applications, such as those related to execution time and power consumption. In particular, many CPUs cannot execute complex object detection algorithms in real time, which is a requirement for automotive ADAS applications and practical Level 3-5 autonomous vehicles.

[0141] In contrast to conventional systems, the techniques described in this paper, by providing CPU complexes, GPU complexes, and hardware accelerator clusters, allow multiple neural networks to be executed simultaneously and / or sequentially, and the results combined to achieve Level 3–5 autonomous driving capabilities. For example, a CNN executed on a DLA or dGPU (e.g., GPU 1120) could include text and word recognition, allowing a supercomputer to read and understand traffic signs, including those for which neural networks have not yet been specifically trained. The DLA could also include a neural network capable of recognizing, interpreting, and providing semantic understanding of the signs, and passing that semantic understanding to a path planning module running on the CPU complex.

[0142] As another example, multiple neural networks can operate simultaneously, as required for Level 3, 4, or 5 driving. For instance, a warning sign consisting of "Caution: Flashing lights indicate icy conditions," along with a light, can be interpreted independently or jointly by several neural networks. The sign itself can be recognized as a traffic sign by a deployed first neural network (e.g., a trained neural network), and the text "Flashing lights indicate icy conditions" can be interpreted by a deployed second neural network that informs the vehicle's path planning software (preferably executing on a CPU complex) that icy conditions exist when the flashing lights are detected. The flashing lights can be identified by a deployed third neural network operating across multiple frames, informing the vehicle's path planning software of the presence (or absence) of the flashing lights. All three neural networks can operate simultaneously, for example, within a DLA and / or on a GPU 1108.

[0143] In some examples, the CNN used for facial recognition and owner identification can use data from camera sensors to identify the presence of an authorized driver and / or owner of vehicle 1100. A processing engine always on the sensors can be used to unlock the vehicle and turn on the lights when the owner approaches the driver's door, and in safe mode, to disable the vehicle when the owner leaves. In this way, SoC 1104 provides security against theft and / or carjacking.

[0144] In another example, the CNN used for emergency vehicle detection and identification can use data from microphone 1196 to detect and identify emergency vehicle siren. In contrast to conventional systems that use a general classifier to detect siren and manually extract features, SoC 1104 uses a CNN to classify environmental and urban sounds as well as visual data. In a preferred embodiment, the CNN running on the DLA is trained to identify the relative shut-off rate of emergency vehicles (e.g., by using the Doppler effect). The CNN can also be trained to identify emergency vehicles specific to the localized area in which the vehicle operates, as identified by GNSS sensor 1158. Thus, for example, when operating in Europe, the CNN will seek to detect European siren, and when operating in the United States, the CNN will seek to identify siren only in North America. Once an emergency vehicle is detected, with the assistance of ultrasonic sensor 1162, the control program can be used to execute emergency vehicle safety routines, causing the vehicle to slow down, pull over to the side of the road, stop, and / or idle until the emergency vehicle passes.

[0145] The vehicle may include a CPU 1118 (e.g., a discrete CPU or dCPU) that can be coupled to the SoC 1104 via a high-speed interconnect (e.g., PCIe). The CPU 1118 may include, for example, an x86 processor. The CPU 1118 can be used to perform any of a wide variety of functions, including, for example, arbitrating the results of potential inconsistencies between ADAS sensors and the SoC 1104, and / or monitoring the status and health of the controller 1136 and / or the infotainment SoC 1130.

[0146] Vehicle 1100 may include a GPU 1120 (e.g., a discrete GPU or dGPU) that can be coupled to SoC 1104 via a high-speed interconnect (e.g., NVIDIA's NVLINK). GPU 1120 may provide additional artificial intelligence capabilities, for example by executing redundant and / or different neural networks, and can be used to train and / or update neural networks based at least in part on inputs (e.g., sensor data) from sensors of vehicle 1100.

[0147] Vehicle 1100 may also include a network interface 1124, which may include one or more wireless antennas 1126 (e.g., one or more wireless antennas for different communication protocols, such as cellular antennas, Bluetooth antennas, etc.). Network interface 1124 can be used to enable wireless connectivity via the Internet to the cloud (e.g., with server 1178 and / or other network devices), with other vehicles, and / or with computing devices (e.g., passenger client devices). For communication with other vehicles, a direct link can be established between the two vehicles, and / or an indirect link can be established (e.g., across networks and via the Internet). A direct link can be provided using a vehicle-to-vehicle communication link. The vehicle-to-vehicle communication link can provide vehicle 1100 with information about vehicles approaching vehicle 1100 (e.g., vehicles in front, to the side, and / or behind vehicle 1100). This functionality may be part of vehicle 1100's cooperative adaptive cruise control function.

[0148] Network interface 1124 may include a SoC that provides modulation and demodulation functions and enables controller 1136 to communicate via a wireless network. Network interface 1124 may include an RF front-end for up-conversion from baseband to RF and down-conversion from RF to baseband. Frequency conversion can be performed using known processes and / or using a superheterodyne process. In some examples, the RF front-end functionality may be provided by a separate chip. The network interface may include wireless functions for communication via LTE, WCDMA, UMTS, GSM, CDMA2000, Bluetooth, Bluetooth LE, Wi-Fi, Z-Wave, ZigBee, LoRaWAN, and / or other wireless protocols.

[0149] Vehicle 1100 may also include data storage 1128, which may include off-chip (e.g., off-chip SoC 1104) storage devices. Data storage 1128 may include one or more storage elements, including RAM, SRAM, DRAM, VRAM, flash memory, hard disk, and / or other components and / or devices capable of storing at least one bit of data.

[0150] Vehicle 1100 may also include a GNSS sensor 1158. The GNSS sensor 1158 (e.g., GPS, assisted GPS sensor, differential GPS (DGPS) sensor, etc.) is used for auxiliary mapping, sensing, occupancy grid generation, and / or path planning functions. Any number of GNSS sensors 1158 can be used, including, for example, but not limited to, GPS using a USB connector with an Ethernet-to-serial (RS-232) bridge.

[0151] Vehicle 1100 may also include a RADAR sensor 1160. The RADAR sensor 1160 can be used by vehicle 1100 for remote vehicle detection even in dark and / or inclement weather conditions. The RADAR functional safety level may be ASIL B. The RADAR sensor 1160 can use CAN and / or bus 1102 (e.g., to transmit data generated by the RADAR sensor 1160) for control and access to object tracking data, and in some examples, Ethernet access for accessing raw data. A wide variety of RADAR sensor types can be used. For example, and without limitation, the RADAR sensor 1160 can be adapted for front, rear, and side RADAR use. In some examples, a pulse Doppler RADAR sensor is used.

[0152] RADAR sensor 1160 can include different configurations, such as long-range with a narrow field of view, short-range with a wide field of view, short-range side coverage, etc. In some examples, long-range RADAR can be used for adaptive cruise control functions. A long-range RADAR system can provide a wide field of view (e.g., within 250m) achieved through two or more independent scans. RADAR sensor 1160 can help distinguish between stationary and moving objects and can be used by ADAS systems for emergency braking assist and forward collision warning. Long-range RADAR sensors can include a single-site multi-mode RADAR with multiple (e.g., six or more) fixed RADAR antennas and high-speed CAN and FlexRay interfaces. In an example with six antennas, the four central antennas can create a focused beam pattern designed to record the vehicle 1100's surroundings at higher rates with minimal traffic interference from adjacent lanes. The other two antennas can extend the field of view, enabling rapid detection of vehicles entering or leaving the vehicle 1100's lane.

[0153] As an example, a mid-range RADAR system can include a range of up to 1160m (front) or 80m (rear) and a field of view of up to 42 degrees (front) or 1150 degrees (rear). Short-range RADAR systems can include, but are not limited to, RADAR sensors designed to be mounted at both ends of the rear bumper. When mounted at both ends of the rear bumper, such a RADAR sensor system can create two beams that continuously monitor blind spots behind and beside the vehicle.

[0154] Short-range RADAR systems can be used in ADAS systems for blind spot detection and / or lane change assistance.

[0155] Vehicle 1100 may also include ultrasonic sensors 1162. Ultrasonic sensors 1162, which may be positioned at the front, rear, and / or sides of vehicle 1100, can be used for parking assistance and / or creating and updating occupancy grids. A wide variety of ultrasonic sensors 1162 can be used, and different ultrasonic sensors 1162 can be used for different detection ranges (e.g., 2.5m, 4m). Ultrasonic sensors 1162 can operate at functional safety level ASIL B.

[0156] Vehicle 1100 may include a LIDAR sensor 1164. The LIDAR sensor 1164 may be used for object and pedestrian detection, emergency braking, collision avoidance, and / or other functions. The LIDAR sensor 1164 may be of functional safety level ASIL B. In some examples, vehicle 1100 may include multiple LIDAR sensors 1164 (e.g., two, four, six, etc.) that can use Ethernet (e.g., to provide data to a Gigabit Ethernet switch).

[0157] In some examples, the LiDAR sensor 1164 may be able to provide a list of objects and their distances within a 360-degree field of view. Commercially available LiDAR sensors 1164 may have an advertising range of, for example, approximately 1100m, with an accuracy of 2cm-3cm, and support for 1100Mbps Ethernet connectivity. In some examples, one or more non-protruding LiDAR sensors 1164 may be used. In such examples, the LiDAR sensor 1164 may be implemented as a small device that can be embedded in the front, rear, sides, and / or corners of the vehicle 1100. In such examples, the LiDAR sensor 1164 may provide a horizontal field of view of up to 120 degrees and a vertical field of view of 35 degrees, even for low-reflectivity objects, with a range of 200m. Front-mounted LiDAR sensors 1164 may be configured for a horizontal field of view between 45 and 135 degrees.

[0158] In some examples, LiDAR technologies such as 3D flash LiDAR can also be used. 3D flash LiDAR uses a flash of laser light as the emission source to illuminate the vehicle's surroundings up to approximately 200 meters. A flash LiDAR unit includes a receiver that records the laser pulse propagation time and reflected light on each pixel, which in turn corresponds to the range from the vehicle to the object. Flash LiDAR allows for the generation of highly accurate and distortion-free images of the surrounding environment using each laser flash. In some examples, four flash LiDAR sensors can be deployed, one on each side of the vehicle. Available 3D flash LiDAR systems include solid-state 3D staring array LiDAR cameras (e.g., non-scanning LiDAR devices) without moving parts other than a fan. Flash LiDAR devices can use 5 nanosecond Class I (eye-safe) laser pulses per frame and can capture reflected laser light in the form of a 3D range point cloud and co-registered intensity data. By using a flash LIDAR, and because a flash LIDAR is a solid-state device with no moving parts, the LIDAR sensor 1164 is less susceptible to motion blur, vibration, and / or shock.

[0159] The vehicle may also include an IMU sensor 1166. In some examples, the IMU sensor 1166 may be located at the center of the rear axle of the vehicle 1100. The IMU sensor 1166 may include, for example, but not limited to, an accelerometer, a magnetometer, a gyroscope, a magnetic compass, and / or other sensor types. In some examples, such as in a six-axis application, the IMU sensor 1166 may include an accelerometer and a gyroscope, while in a nine-axis application, the IMU sensor 1166 may include an accelerometer, a gyroscope, and a magnetometer.

[0160] In some embodiments, the IMU sensor 1166 can be implemented as a miniature, high-performance GPS-assisted inertial navigation system (GPS / INS) that combines a microelectromechanical system (MEMS) inertial sensor, a high-sensitivity GPS receiver, and an advanced Kalman filter algorithm to provide estimates of position, velocity, and attitude. Thus, in some examples, the IMU sensor 1166 can enable the vehicle 1100 to estimate heading by directly observing and correlating velocity changes from GPS to the IMU sensor 1166 without requiring input from a magnetic sensor. In some examples, the IMU sensor 1166 and the GNSS sensor 1158 can be combined into a single integrated unit.

[0161] The vehicle may include a microphone 1196 placed in and / or around the vehicle 1100. Among other things, the microphone 1196 may be used for emergency vehicle detection and identification.

[0162] The vehicle may also include any number of camera types, including stereo camera 1168, wide-angle camera 1170, infrared camera 1172, surround camera 1174, long-range and / or mid-range camera 1198, and / or other camera types. These cameras can be used to capture image data around the entire perimeter of the vehicle 1100. The types of cameras used depend on the embodiment and the requirements of the vehicle 1100, and any combination of camera types can be used to provide the necessary coverage around the vehicle 1100. Furthermore, the number of cameras may vary depending on the embodiment. For example, the vehicle may include six cameras, seven cameras, ten cameras, twelve cameras, and / or another number of cameras. As an example and without limitation, these cameras may support Gigabit Multimedia Serial Link (GMSL) and / or Gigabit Ethernet. Each of the cameras is described herein with respect to... Figure 11A and Figure 11B It was described in more detail.

[0163] Vehicle 1100 may also include vibration sensor 1142. Vibration sensor 1142 can measure vibrations of vehicle components such as axles. For example, changes in vibration can indicate changes in the road surface. In another example, when two or more vibration sensors 1142 are used, differences between vibrations can be used to determine friction or slippage on the road surface (e.g., when there is a vibration difference between a power drive shaft and a free-rotating shaft).

[0164] Vehicle 1100 may include ADAS system 1138. In some examples, ADAS system 1138 may include SoC. ADAS system 1138 may include autonomous / adaptive / automatic cruise control (ACC), cooperative adaptive cruise control (CACC), forward collision warning (FCW), automatic emergency braking (AEB), lane departure warning (LDW), lane keeping assist (LKA), blind spot warning (BSW), rear cross traffic warning (RCTW), collision warning system (CWS), lane centering (LC) and / or other features and functions.

[0165] The ACC system can use RADAR sensor 1160, LIDAR sensor 1164, and / or a camera. The ACC system may include longitudinal ACC and / or lateral ACC. Longitudinal ACC monitors and controls the distance to vehicles immediately in front of vehicle 1100 and automatically adjusts the vehicle speed to maintain a safe distance. Lateral ACC performs distance holding and, if necessary, advises vehicle 1100 to change lanes. Lateral ACC is associated with other ADAS applications such as LCA and CWS.

[0166] CACC uses information from other vehicles, which can be received indirectly from other vehicles via a wireless link or through a network connection (e.g., via the Internet) through network interface 1124 and / or wireless antenna 1126. Direct links can be provided by vehicle-to-vehicle (V2V) communication links, while indirect links can be infrastructure-to-vehicle (I2V) communication links. Typically, the V2V communication concept provides information about vehicles immediately ahead (e.g., vehicles immediately in front of vehicle 1100 and in the same lane), while the I2V communication concept provides information about traffic further ahead. A CACC system can include either or both of these I2V and V2V information sources. Given information about vehicles ahead of vehicle 1100, CACC can be more reliable, and it has the potential to improve traffic flow and reduce road congestion.

[0167] The Forward-Looking Warning (FCW) system is designed to alert the driver to hazards, enabling the driver to take corrective action. The FCW system uses a front-facing camera and / or RADAR sensor 1160 coupled to a dedicated processor, DSP, FPGA, and / or ASIC, which is electrically coupled to driver feedback such as a display, speaker, and / or vibrating components. The FCW system can provide warnings in the form of, for example, audible, visual, haptic, and / or rapid braking pulses.

[0168] The AEB system detects an impending forward collision with another vehicle or other object and can automatically apply the brakes if the driver does not take corrective action within a specified time or distance parameter. The AEB system can use a front-facing camera and / or RADAR sensor 1160 coupled to a dedicated processor, DSP, FPGA, and / or ASIC. When the AEB system detects a hazard, it typically first alerts the driver to take corrective action to avoid a collision, and if the driver does not take corrective action, the AEB system can automatically apply the brakes to attempt to prevent or at least mitigate the effects of the predicted collision. The AEB system may include technologies such as dynamic brake support and / or collision proximity braking.

[0169] The Lane Departure Warning (LDW) system provides visual, auditory, and / or tactile warnings, such as steering wheel or seat vibrations, to alert the driver when the vehicle crosses lane markings. When the driver indicates intentional lane departure, the LDW system is deactivated by activating a turn signal. The LDW system can utilize a front-facing camera coupled to a dedicated processor, DSP, FPGA, and / or ASIC, which is electrically coupled to driver feedback such as a display, speaker, and / or vibrating components.

[0170] The LKA system is a variant of the LDW system. If vehicle 1100 begins to leave the lane, the LKA system provides steering input or braking to correct vehicle 1100.

[0171] The BSW system detects and warns the driver of vehicles in the vehicle's blind spot. The BSW system can provide visual, auditory, and / or tactile alerts to indicate that merging or changing lanes is unsafe. The system can provide additional warnings when the driver uses turn signals. The BSW system can use a rear-facing camera and / or RADAR sensor 1160 coupled to a dedicated processor, DSP, FPGA, and / or ASIC, which is electrically coupled to driver feedback such as a display, speaker, and / or vibrating components.

[0172] The RCTW system can provide visual, auditory, and / or tactile notifications when an object is detected outside the range of the rear camera while the vehicle 1100 is reversing. Some RCTW systems include AEB to ensure the application of the vehicle's brakes to avoid a collision. The RCTW system may use one or more rear-mounted RADAR sensors 1160 coupled to a dedicated processor, DSP, FPGA, and / or ASIC, which is electrically coupled to driver feedback such as a display, speaker, and / or vibrating components.

[0173] Conventional ADAS systems can be prone to false positives, which can be annoying and distracting for the driver, but typically not catastrophic, as they alert the driver and allow them to determine whether a safe condition truly exists and take appropriate action. However, in the autonomous vehicle 1100, in the event of conflicting results, the vehicle 1100 itself must decide whether to heed the results from the main computer or auxiliary computer (e.g., the first controller 1136 or the second controller 1136). For example, in some embodiments, the ADAS system 1138 may be a backup and / or auxiliary computer for providing perception information to a backup computer rationality module. The backup computer rationality monitor may run redundant and varied software on hardware components to detect faults in perception and dynamic driving tasks. Outputs from the ADAS system 1138 may be provided to a supervisory MCU. If the outputs from the main computer and the auxiliary computer conflict, the supervisory MCU must determine how to reconcile the conflict to ensure safe operation.

[0174] In some examples, the master computer can be configured to provide a confidence score to the supervisory MCU, indicating the master computer's confidence level in the selected result. If the confidence score exceeds a threshold, the supervisory MCU can follow the master computer's direction regardless of whether the auxiliary computer provides conflicting or inconsistent results. If the confidence score does not meet the threshold and the master and auxiliary computers indicate different results (e.g., conflict), the supervisory MCU can arbitrate between these computers to determine the appropriate result.

[0175] The supervisory MCU can be configured to run a neural network trained and configured to determine the conditions under which the auxiliary computer provides a false alarm, based at least in part on outputs from both the host and auxiliary computers. Thus, the neural network in the supervisory MCU can learn when the output of the auxiliary computer can be trusted and when it cannot. For example, when the auxiliary computer is a RADAR-based FCW system, the neural network in the supervisory MCU can learn when the FCW system is identifying a metallic object that is not actually dangerous, such as a drain grid or manhole cover that triggers an alarm. Similarly, when the auxiliary computer is a camera-based LDW system, the neural network in the supervisory MCU can learn to ignore the LDW when a cyclist or pedestrian is present and lane departure is actually the safest strategy. In embodiments that include a neural network running on the supervisory MCU, the supervisory MCU may include at least one of a DLA or GPU suitable for running the neural network using associated memory. In a preferred embodiment, the supervisory MCU may include a component of SoC 1104 and / or be included as a component of SoC 1104.

[0176] In other examples, ADAS system 1138 may include an auxiliary computer that performs ADAS functions using conventional computer vision rules. This allows the auxiliary computer to use classic computer vision rules (if-then), and the presence of neural networks in the supervising MCU can improve reliability, safety, and performance. For example, diverse implementations and intentional non-identity make the entire system more fault-tolerant, especially for failures caused by software (or software-hardware interface) functionality. For instance, if a software vulnerability or bug exists in the software running on the host computer and non-identical software code running on the auxiliary computer provides the same overall result, the supervising MCU can be more confident that the overall result is correct and that the vulnerability in the software or hardware on the host computer does not cause a substantial error.

[0177] In some examples, the output of ADAS system 1138 can be fed to the perception block and / or the dynamic driving task block of the main computer. For example, if ADAS system 1138 issues a forward collision warning because an object is immediately in front, the perception block can use this information when identifying the object. In other examples, the assistance computer can have its own neural network, which is trained and thus reduces the risk of false positives as described herein.

[0178] Vehicle 1100 may also include an infotainment SoC 1130 (e.g., an in-vehicle infotainment system (IVI)). Although illustrated and described as an SoC, the infotainment system may not be an SoC and may include two or more discrete components. The infotainment SoC 1130 may include a combination of hardware and software that can be used to provide vehicle 1100 with audio (e.g., music, personal digital assistant, navigation instructions, news, radio, etc.), video (e.g., TV, movies, streaming media, etc.), telephone (e.g., hands-free calling), network connectivity (e.g., LTE, WiFi, etc.) and / or information services (e.g., navigation system, rear parking assistance, radio data system, vehicle-related information such as fuel level, total coverage distance, brake fuel level, fuel level, door opening / closing, air filter information, etc.). For example, the infotainment SoC 1130 may include a radio, disc player, navigation system, video player, USB and Bluetooth connectivity, in-vehicle computer, in-vehicle entertainment, WiFi, steering wheel audio controls, hands-free voice controls, head-up display (HUD), HMI display 1134, telematics device, control panel (e.g., for controlling and / or interacting with various components, features, and / or systems) and / or other components. The infotainment SoC 1130 may further be used to provide information (e.g., visual and / or auditory) to users of the vehicle, such as information from the ADAS system 1138, autonomous driving information such as planned vehicle maneuvers, trajectories, surrounding environment information (e.g., intersection information, vehicle information, road information, etc.), and / or other information.

[0179] The infotainment SoC 1130 may include GPU functionality. The infotainment SoC 1130 can communicate with other devices, systems, and / or components of the vehicle 1100 via bus 1102 (e.g., CAN bus, Ethernet, etc.). In some examples, the infotainment SoC 1130 may be coupled to a supervisory MCU, allowing the GPU of the infotainment system to perform some autonomous driving functions in the event of a failure of the main controller 1136 (e.g., the primary and / or backup computer of the vehicle 1100). In such an example, the infotainment SoC 1130 may place the vehicle 1100 into a driver-safe parking mode as described herein.

[0180] Vehicle 1100 may also include instrument cluster 1132 (e.g., digital instrument panel, electronic instrument cluster, digital instrument panel, etc.). Instrument cluster 1132 may include a controller and / or supercomputer (e.g., a discrete controller or supercomputer). Instrument cluster 1132 may include a set of instruments such as speedometer, fuel level, oil pressure, tachometer, odometer, turn indicator, shift position indicator, seatbelt warning light, parking brake warning light, engine malfunction indicator, airbag (SRS) system information, lighting controls, safety system controls, navigation information, etc. In some examples, information may be displayed and / or shared between infotainment SoC 1130 and instrument cluster 1132. In other words, instrument cluster 1132 may be included as part of infotainment SoC 1130, or vice versa.

[0181] Figure 11D For cloud-based servers and according to some embodiments of this disclosure Figure 11A The following is a system diagram illustrating communication between example autonomous vehicles 1100. System 1176 may include server 1178, network 1190, and vehicles including vehicle 1100. Server 1178 may include multiple GPUs 1184(A)-1284(H) (collectively referred to herein as GPU 1184), PCIe switches 1182(A)-1182(H) (collectively referred to herein as PCIe switch 1182), and / or CPUs 1180(A)-1180(B) (collectively referred to herein as CPU 1180). GPU 1184, CPU 1180, and PCIe switches may be interconnected with high-speed interconnects and / or PCIe connections 1186, such as, but not limited to, NVLink interface 1188 developed by NVIDIA. In some examples, GPU 1184 is connected via NVLink and / or NVSwitch SoC, and GPU 1184 and PCIe switch 1182 are connected via PCIe interconnect. Although the diagram illustrates eight GPUs 1184, two CPUs 1180, and two PCIe switches, it is not intended to be limiting. Depending on the embodiment, each of the servers 1178 may include any number of GPUs 1184, CPUs 1180, and / or PCIe switches. For example, each of the servers 1178 may include eight, sixteen, thirty-two, and / or more GPUs 1184.

[0182] Server 1178 can receive image data from vehicles via network 1190, representing images of unexpected or altered road conditions such as recently commenced roadworks. Server 1178 can also transmit neural network 1192, updated neural network 1192, and / or map information 1194, including information about traffic and road conditions, to vehicles via network 1190. Updates to map information 1194 may include updates to HD map 1122, such as information about construction sites, potholes, bends, floods, or other obstacles. In some examples, neural network 1192, updated neural network 1192, and / or map information 1194 may have been generated from new training and / or data received from any number of vehicles in the environment, and / or based on experience gained from training performed at a data center (e.g., using server 1178 and / or other servers).

[0183] Server 1178 can be used to train machine learning models (e.g., neural networks) based on training data. Training data can be generated by the vehicle and / or generated in a simulation (e.g., using a game engine). In some examples, the training data is labeled (e.g., where the neural network benefits from supervised learning) and / or undergoes other preprocessing, while in other examples, the training data is not labeled and / or preprocessed (e.g., where the neural network does not require supervised learning). Training can be performed according to any one or more classes of machine learning techniques, including but not limited to: supervised training, semi-supervised training, unsupervised training, self-learning, reinforcement learning, joint learning, transfer learning, feature learning (including principal component and cluster analysis), multilinear subspace learning, manifold learning, representation learning (including alternative dictionary learning), rule-based machine learning, anomaly detection, and any variations or combinations thereof. Once the machine learning model is trained, it can be used by the vehicle (e.g., transmitted to the vehicle via network 1190), and / or the machine learning model can be used by server 1178 to remotely monitor the vehicle.

[0184] In some examples, server 1178 can receive data from a vehicle and apply that data to a state-of-the-art real-time neural network for real-time intelligent inference. Server 1178 may include a deep learning supercomputer powered by GPU 1184 and / or a dedicated AI computer, such as the DGX and DGX Station machines developed by NVIDIA. However, in some examples, server 1178 may include a deep learning infrastructure in a data center that uses only CPU power.

[0185] The deep learning infrastructure of server 1178 may be capable of rapid real-time inference and can be used to assess and verify the health status of the processor, software, and / or associated hardware in vehicle 1100. For example, the deep learning infrastructure may receive periodic updates from vehicle 1100, such as image sequences and / or objects located in those image sequences that vehicle 1100 has already located (e.g., via computer vision and / or other machine learning object classification techniques). The deep learning infrastructure may run its own neural network to identify objects and compare them with objects identified by vehicle 1100. If the results do not match and the infrastructure concludes that the AI ​​in vehicle 1100 has malfunctioned, then server 1178 may transmit a signal to vehicle 1100 instructing the vehicle's fail-safe computer to take control, notify passengers, and complete a safe stopping operation.

[0186] For inference, server 1178 may include GPU 1184 and one or more programmable inference accelerators (such as NVIDIA's TensorRT). The combination of a GPU-powered server and inference acceleration enables real-time response. In other examples, such as where performance is less critical, CPU, FPGA, and other processor-powered servers can be used for inference.

[0187] Example computing device

[0188] Figure 12 This is a block diagram of an example computing device 1200 suitable for implementing some embodiments of the present disclosure. The computing device 1200 may include an interconnect system 1202 directly or indirectly coupled to the following devices: a memory 1204, one or more central processing units (CPUs) 1206, one or more graphics processing units (GPUs) 1208, a communication interface 1210, input / output (I / O) ports 1212, input / output components 1214, a power supply 1216, one or more presentation components 1218 (e.g., one or more displays), and one or more logic units 1220. In at least one embodiment, the computing device 1200 may include one or more virtual machines (VMs), and / or any component thereof may include virtual components (e.g., virtual hardware components). For a non-limiting example, one or more GPUs 1208 may include one or more vGPUs, one or more CPUs 1206 may include one or more vCPUs, and / or one or more logic units 1220 may include one or more virtual logic units. Thus, (one or more) computing devices 1200 may include discrete components (e.g., a full GPU dedicated to computing device 1200), virtual components (e.g., a portion of the GPU dedicated to computing device 1200), or a combination thereof.

[0189] although Figure 12 The various blocks are shown as connected via interconnect system 1202 using lines, but this is not intended to be limiting and is merely for clarity. For example, in some embodiments, presentation component 1218 (such as a display device) may be considered I / O component 1214 (e.g., if the display is a touchscreen). As another example, CPU 1206 and / or GPU 1208 may include memory (e.g., memory 1204 may represent a storage device other than the memory of GPU 1208, CPU 1206, and / or other components). In other words, Figure 12 The computing devices described are for illustrative purposes only. No distinction is made between such categories as “workstation,” “server,” “laptop computer,” “desktop computer,” “tablet computer,” “client device,” “mobile device,” “handheld device,” “game console,” “electronic control unit (ECU),” “virtual reality system,” and / or other device or system types, as all are considered within the scope of… Figure 12 Within the scope of computing devices.

[0190] Interconnect system 1202 may represent one or more links or buses, such as address buses, data buses, control buses, or combinations thereof. Interconnect system 1202 may include one or more bus or link types, such as Industry Standard Architecture (ISA) bus, Extended Industry Standard Architecture (EISA) bus, Video Electronics Standards Association (VESA) bus, Peripheral Component Interconnect (PCI) bus, Fast Peripheral Component Interconnect (PCIe) bus, and / or another type of bus or link. In some embodiments, there is a direct connection between components. As an example, CPU 1206 may be directly connected to memory 1204. Further, CPU 1206 may be directly connected to GPU 1208. In cases where there is a direct or point-to-point connection between components, interconnect system 1202 may include a PCIe link to perform the connection. In these examples, a PCI bus is not required to be included in computing device 1200.

[0191] The memory 1204 may include any computer-readable medium from a variety of computer-readable media. The computer-readable medium may be any available medium accessible by the computing device 1200. The computer-readable medium may include volatile and non-volatile media, as well as removable and non-removable media. By way of example and not limitation, the computer-readable medium may include computer storage media and communication media.

[0192] Computer storage media may include volatile and non-volatile media and / or removable and non-removable media implemented with any method or technology for storing information such as computer-readable instructions, data structures, program modules and / or other data types. For example, memory 1204 may store computer-readable instructions (e.g., representing (one or more) programs and / or (one or more) program elements, such as an operating system). Computer storage media may include, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic tape cassettes, magnetic tape, disk storage devices or other magnetic storage devices, or any other medium that can be used to store the desired information and is accessible by computing device 1200. As used herein, computer storage media does not include the signal itself.

[0193] Computer storage media can embody computer-readable instructions, data structures, program modules, and / or other data types in modulated data signals such as carrier waves or other transmission mechanisms, and includes any information transmission medium. The term "modulated data signal" can refer to a signal whose one or more characteristics are set or altered in a manner that encodes information in the signal. By way of example and not limitation, computer storage media can include wired media (such as wired networks or direct wired connections) and wireless media (such as acoustic, RF, infrared, and other wireless media). Any combination of the above should also be included within the scope of computer-readable media.

[0194] CPU 1206 may be configured to execute at least some of computer-readable instructions to control one or more components of computing device 1200 to perform one or more of the methods and / or processes described herein. Each CPU 1206 may contain one or more cores (e.g., one, two, four, eight, twenty-eight, seventy-two, etc.) capable of handling numerous software threads simultaneously. CPU 1206 may contain any type of processor and may contain different types of processors depending on the type of computing device 1200 implemented (e.g., processors with fewer cores for mobile devices and processors with more cores for servers). For example, depending on the type of computing device 1200, the processor may be an advanced RISC machine (ARM) processor implemented using Reduced Instruction Set Computing (RISC) or an x86 processor implemented using Complex Instruction Set Computing (CISC). In addition to one or more microprocessors or supplementary coprocessors (such as math coprocessors), computing device 1200 may also include one or more CPUs 1206.

[0195] In addition to or in lieu of one or more CPUs 1206, one or more GPUs 1208 may be configured to execute at least some of computer-readable instructions to control one or more components of computing device 1200 to perform one or more of the methods and / or processes described herein. One or more GPUs 1208 may be integrated GPUs (e.g., with one or more CPUs 1206) and / or one or more GPUs 1208 may be discrete GPUs. In embodiments, one or more GPUs 1208 may be coprocessors of one or more CPUs 1206. GPUs 1208 may be used by computing device 1200 to render graphics (e.g., 3D graphics) or perform general-purpose computing. For example, GPUs 1208 may be used for general-purpose computing on a GPU (GPGPU). GPUs 1208 may contain hundreds or thousands of cores capable of handling hundreds or thousands of software threads simultaneously. GPU 1208 can generate pixel data for an output image in response to rendering commands (e.g., rendering commands received from CPU 1206 via a host interface). GPU 1208 may include graphics memory (e.g., display memory) for storing pixel data or any other suitable data (e.g., GPGPU data). Display memory may be included as part of memory 1204. GPU 1208 may include two or more GPUs operating in parallel (e.g., via links). The links may directly connect the GPUs (e.g., using NVLINK) or may connect the GPUs via a switch (e.g., using NVSwitch). When combined, each GPU 1208 may generate pixel data or GPGPU data for different portions of the output or for different outputs (e.g., a first GPU for a first image and a second GPU for a second image). Each GPU may include its own memory or may share memory with other GPUs.

[0196] In addition to or in lieu of CPU 1206 and / or GPU 1208, logic unit 1220 may be configured to execute at least some of computer-readable instructions to control one or more components of computing device 1200 to perform one or more of the methods and / or processes described herein. In embodiments, one or more CPUs 1206, one or more GPUs 1208, and / or one or more logic units 1220 may perform any combination of methods, processes, and / or portions thereof discretely or jointly. One or more logic units 1220 may be a portion of one or more CPUs 1206 and / or GPUs 1208 and / or integrated into one or more CPUs 1206 and / or GPUs 1208, and / or one or more logic units 1220 may be discrete components or otherwise external to CPUs 1206 and / or GPUs 1208. In an embodiment, one or more of the logic units 1220 may be coprocessors of one or more of the CPU 1206 and / or one or more of the GPU 1208.

[0197] Examples of logic unit 1220 include one or more processing cores and / or components thereof, such as data processing unit (DPU), tensor core (TC), tensor processing unit (TPU), pixel vision core (PVC), vision processing unit (VPU), graphics processing cluster (GPC), texture processing cluster (TPC), streaming multiprocessor (SM), tree lateral unit (TTU), artificial intelligence accelerator (AIA), deep learning accelerator (DLA), arithmetic logic unit (ALU), application-specific integrated circuit (ASIC), floating-point unit (FPU), input / output (I / O) element, peripheral component interconnect (PCI) or fast peripheral component interconnect (PCIe) element, etc.

[0198] Communication interface 1210 may include one or more receivers, transmitters, and / or transceivers enabling computing device 1200 to communicate with other computing devices via electronic communication networks (including wired and / or wireless communications). Communication interface 1210 may include components and functions for enabling communication over any of a plurality of different networks, such as wireless networks (e.g., Wi-Fi, Z-Wave, Bluetooth, Bluetooth LE, ZigBee, etc.), wired networks (e.g., via Ethernet or wirelessband communication), low-power wide area networks (e.g., LoRaWAN, SigFox, etc.), and / or the Internet. In one or more embodiments, one or more logic units 1220 and / or communication interface 1210 may include one or more data processing units (DPUs) for directly transmitting data received via a network and / or via interconnect system 1202 to one or more GPUs 1208 (e.g., their memory).

[0199] I / O port 1212 enables computing device 1200 to be logically coupled to other devices including I / O component 1214, (one or more) presentation component 1218, and / or other components, some of which may be built into (e.g., integrated into) computing device 1200. Illustrative I / O component 1214 includes microphones, mice, keyboards, joysticks, gamepads, game controllers, satellite dish antennas, scanners, printers, wireless devices, etc. I / O component 1214 can provide a natural user interface (NUI) that processes aerial gestures, voice, or other physiological input generated by the user. In some cases, input may be transmitted to appropriate network elements for further processing. The NUI can implement any combination of voice recognition, pen recognition, facial recognition, biometric recognition, on-screen and near-screen gesture recognition, aerial gestures, head and eye tracking, and touch recognition (as described in more detail below) associated with the display of computing device 1200. Computing device 1200 may include depth cameras for gesture detection and recognition, such as stereo camera systems, infrared camera systems, RGB camera systems, touchscreen technology, and combinations thereof. Additionally, the computing device 1200 may include an accelerometer or gyroscope (e.g., as part of an inertial measurement unit (IMU)) that enables motion detection. In some examples, the computing device 1200 may use the output of the accelerometer or gyroscope to render immersive augmented reality or virtual reality.

[0200] Power supply 1216 may include a hardwired power supply, a battery power supply, or a combination thereof. Power supply 1216 may provide power to computing device 1200 to enable the components of computing device 1200 to operate.

[0201] The presentation component 1218 may include a display (e.g., a monitor, touchscreen, television screen, head-up display (HUD), other display types, or combinations thereof), speakers, and / or other presentation components. The presentation component 1218 may receive data from other components (e.g., GPU 1208, CPU 1206, DPU, etc.) and output the data (e.g., as images, videos, sounds, etc.).

[0202] Example Data Center

[0203] Figure 13 An example data center 1300 that may be used in at least one embodiment of this disclosure is shown. The data center 1300 may include a data center infrastructure layer 1310, a framework layer 1320, a software layer 1330, and / or an application layer 1340.

[0204] like Figure 13 As shown, the data center infrastructure layer 1310 may include a resource coordinator 1312, grouped computing resources 1314, and node computing resources (“nodes CRs”) 1316(1)-1316(N), where “N” represents any complete positive integer. In at least one embodiment, nodes CRs 1316(1)-1316(N) may include, but are not limited to, any number of central processing units (CPUs) or other processors (including DPUs, accelerators, field-programmable gate arrays (FPGAs), graphics processing units or graphics processing units (GPUs), etc.), memory devices (e.g., dynamic read-only memory), storage devices (e.g., solid-state or disk drives), network input / output (NW I / O) devices, network switches, virtual machines (VMs), power modules and / or cooling modules, etc. In some embodiments, one or more node CRs from nodes CRs 1316(1)-1316(N) may correspond to servers having one or more of the aforementioned computing resources. In addition, in some embodiments, nodes CRs1316(1)-13161(N) may include one or more virtual components, such as vGPU, vCPU, etc., and / or one or more nodes CRs1316(1)-1316(N) may correspond to virtual machines (VMs).

[0205] In at least one embodiment, the grouped computing resources 1314 may include individual groups of node CRs 1316 housed within one or more racks (not shown), or a plurality of racks housed within a data center in different geographical locations (also not shown). Individual groups of node CRs 1316 within the grouped computing resources 1314 may include grouped computing, networking, memory, or storage resources that can be configured or allocated to support one or more workloads. In at least one embodiment, a plurality of node CRs 1316, including CPUs, GPUs, DPUs, and / or other processors, may be grouped within one or more racks to provide computing resources to support one or more workloads. One or more racks may also include any number of power modules, cooling modules, and / or network switches in any combination.

[0206] Resource coordinator 1312 may be configured or otherwise control one or more nodes CRs 1316(1)-1316(N) and / or grouped computing resources 1314. In at least one embodiment, resource coordinator 1312 may include a Software Design Infrastructure (SDI) management entity for data center 1300. Resource coordinator 1312 may include hardware, software, or some combination thereof.

[0207] In at least one embodiment, such as Figure 13As shown, framework layer 1320 may include job scheduler 1333, configuration manager 1334, resource manager 1336, and / or distributed file system 1338. Framework layer 1320 may include a framework of software 1332 supporting software layer 1330 and / or one or more applications 1342 supporting application layer 1340. Software 1332 or application 1342 may respectively contain web-based service software or applications, such as those provided by Amazon Web Services, Google Cloud, and Microsoft Azure. Framework layer 1320 may be, but is not limited to, a free and open-source software web application framework (such as Apache Spark™ (hereinafter “Spark”)) that can leverage distributed file system 1338 for large-scale data processing (e.g., “big data”). In at least one embodiment, job scheduler 1333 may include Spark drivers to facilitate the scheduling of workloads supported by different layers of data center 1300. Configuration manager 1334 may be able to configure different layers, such as software layer 1330 and framework layer 1320 (which includes Spark and distributed file system 1338 for supporting large-scale data processing). Resource manager 1336 may be able to manage compute resources mapped to or allocated to clusters of distributed file system 1338 and job scheduler 1333, or allocated to support clusters of distributed file system 1338 and job scheduler 1333. In at least one embodiment, clustered or grouped compute resources may include grouped compute resources 1314 in data center infrastructure layer 1310. Resource manager 1336 may coordinate with resource coordinator 1312 to manage these mapped or allocated compute resources.

[0208] In at least one embodiment, the software 1332 included in software layer 1330 may include software used in at least a portion of the nodes CRs 1316(1)-1316(N), the grouped computing resources 1314, and / or the distributed file system 1338 of framework layer 1320. One or more types of software may include, but are not limited to, internet web search software, email virus scanning software, database software, and streaming video content software.

[0209] In at least one embodiment, the application 1342 included in the application layer 1340 may include one or more types of applications used at least in part by nodes CRs 1316(1)-1316(N), grouped computing resources 1314, and / or the distributed file system 1338 of the framework layer 1320. One or more types of applications may include, but are not limited to, any number of genomics applications, cognitive computing, and machine learning applications, including training or inference software, machine learning framework software (e.g., PyTorch, TensorFlow, Caffe, etc.), and / or other machine learning applications used in combination with one or more embodiments.

[0210] In at least one embodiment, any of the configuration manager 1334, resource manager 1336, and resource coordinator 1312 can implement any number and type of self-modification actions based on any amount and type of data acquired in any technically feasible manner. Self-modification actions can free data center operators of data center 1300 from making potentially poor configuration decisions and may prevent underutilization and / or poor performance of the data center.

[0211] According to one or more embodiments described herein, data center 1300 may include tools, services, software, or other resources to train one or more machine learning models or to use one or more machine learning models to predict or infer information. For example, one or more machine learning models can be trained by using the software and / or computing resources described above with respect to data center 1300 to compute weight parameters according to a neural network architecture. In at least one embodiment, a trained or deployed machine learning model corresponding to one or more neural networks can be used to infer or predict information using the resources described above with respect to data center 1300 by using weight parameters computed through one or more training techniques (such as, but not limited to, those described herein).

[0212] In at least one embodiment, the data center 1300 may use a CPU, application-specific integrated circuit (ASIC), GPU, FPGA, and / or other hardware (or corresponding virtual computing resources) to perform training and / or inference using the aforementioned resources. Furthermore, one or more of the software and / or hardware resources described above may be configured to allow a user to train or perform services that infer information, such as image recognition, speech recognition, or other artificial intelligence services.

[0213] Example network environment

[0214] A network environment suitable for implementing embodiments of this disclosure may include one or more client devices, servers, network-attached storage (NAS), other backend devices, and / or other device types. Client devices, servers, and / or other device types (e.g., each device) may be... Figure 12 This is implemented on one or more instances of computing device 1200—for example, each device may include similar components, features, and / or functions of computing device 1200. Furthermore, in the case of implementing backend devices (e.g., servers, NAS, etc.), the backend devices may be included as part of data center 1300, examples of which are described herein. Figure 13 To describe in more detail.

[0215] Components of a network environment can communicate with each other via a network, which can be wired, wireless, or both. A network can include multiple networks or one of multiple networks. For example, a network can include one or more wide area networks (WANs), one or more local area networks (LANs), one or more public networks (such as the Internet and / or the Public Switched Telephone Network (PSTN)), and / or one or more private networks. Where the network includes a wireless telecommunications network, components such as base stations, communication towers, or even access points (and other components) can provide wireless connectivity.

[0216] A compatible network environment may include one or more peer-to-peer network environments (in which case the server may not be included in the network environment) and one or more client-server network environments (in which case one or more servers may be included in the network environment). In a peer-to-peer network environment, the functionality described herein for the server can be implemented on any number of client devices.

[0217] In at least one embodiment, the network environment may include one or more cloud-based network environments, distributed computing environments, combinations thereof, etc. The cloud-based network environment may include a framework layer, a job scheduler, a resource manager, and a distributed file system implemented on one or more servers, which may include one or more core network servers and / or edge servers. The framework layer may include a framework supporting software at the software layer and / or application at the application layer. The software or application may respectively include network-based service software or applications. In embodiments, one or more client devices may use the network-based service software or applications (e.g., by accessing the service software and / or applications via one or more application programming interfaces (APIs)). The framework layer may be, but is not limited to, a free and open-source software network application framework that can use a distributed file system for large-scale data processing (e.g., "big data").

[0218] A cloud-based network environment can provide cloud computing and / or cloud storage for any combination of the computing and / or data storage functions (or one or more portions thereof) described herein. Any of these different functions can be distributed across multiple locations from a central or core server (e.g., distributed across one or more data centers at the state, region, country, global, etc.). The core server may assign at least a portion of the functionality to the edge server if the connection to the user (e.g., a client device) is relatively close to the edge server. A cloud-based network environment can be private (e.g., limited to a single organization), public (e.g., available to many organizations), and / or a combination thereof (e.g., a hybrid cloud environment).

[0219] (One or more) client devices may include the information described in this article. Figure 12 At least some of the components, features, and functions of the described (one or more) example computing device 1200. By way of example and not limitation, the client device may be implemented as a personal computer (PC), laptop computer, mobile device, smartphone, tablet computer, smartwatch, wearable computer, personal digital assistant (PDA), MP3 player, virtual reality headset, global positioning system (GPS) or device, video player, camera, surveillance equipment or system, vehicle, ship, spacecraft, virtual machine, drone, robot, handheld communication device, hospital equipment, gaming equipment or system, entertainment system, vehicle computer system, embedded system controller, remote control, electrical appliance, consumer electronics device, workstation, edge device, any combination of these depicted devices, or any other suitable device.

[0220] This disclosure can be described in the general context of machine-usable instructions or computer code, including computer-executable instructions such as program modules, which are executed by a computer or other machine such as a personal digital assistant or other handheld device. Typically, a program module, including routines, programs, objects, components, data structures, etc., refers to code that performs a specific task or implements a specific abstract data type. This disclosure can be practiced in a wide variety of system configurations, including handheld devices, consumer electronics, general-purpose computers, more specialized computing devices, etc. This disclosure can also be practiced in distributed computing environments where tasks are performed by remote processing devices linked via a communication network.

[0221] As used herein, the phrase "and / or" relating to two or more elements should be interpreted as referring to only one element or a combination of elements. For example, "element A, element B, and / or element C" can include only element A, only element B, only element C, element A and element B, element A and element C, element B and element C, or element A, B, and C. Furthermore, "at least one of element A or element B" can include at least one of element A, at least one of element B, or at least one of element A and at least one of element B. Further, "at least one of element A and element B" can include at least one of element A, at least one of element B, or at least one of element A and at least one of element B.

[0222] This document describes in detail the subject matter of this disclosure to meet statutory requirements. However, the description itself is not intended to limit the scope of this disclosure. Rather, the discloser has envisioned that the claimed subject matter may be embodied in other ways to include steps different from or similar combinations of steps described herein in conjunction with other current or future techniques. Moreover, although the terms “step” and / or “box” may be used herein to imply different elements of the method employed, these terms should not be construed as suggesting any particular order among or between the various steps disclosed herein, unless the order of the steps is explicitly described.

[0223] Example paragraph

[0224] A: A method comprising: using perception data and based at least on lateral distances between one or more first lanes and one or more second lanes, determining one or more first locations in an environment corresponding to one or more first merging points, at which the one or more first lanes deviate from the one or more second lanes; determining one or more second locations in the environment corresponding to one or more second merging points, at least based on map data representing a map of the environment, at which one or more first road segments deviate from one or more second road segments; at least using the one or more first locations and the one or more second locations, associating the one or more first lanes with the one or more first road segments and associating the one or more second lanes with the one or more second road segments; and performing one or more operations associated with a machine in the environment, at least based on the association.

[0225] B. The method as described in paragraph A further includes: preprocessing the perceived data to remove one or more features from the perceived data, the one or more features including at least one of the following: one or more shoulder lanes; one or more oncoming lanes; one or more bicycle lanes; or one or more portions of the one or more first lanes or the one or more second lanes having a variance that meets or exceeds a threshold.

[0226] C. The method as described in any one of paragraphs A and B, wherein determining the one or more first locations in the environment corresponding to the one or more first merging points comprises: detecting, at least based on the perception data, that a second lateral distance between the one or more first lanes and the one or more second lanes satisfies or exceeds a first threshold after the one or more first merging points; and determining the one or more first locations corresponding to the one or more first merging points, at least based on the lateral distance being less than a second threshold near the one or more first locations.

[0227] D. The method as described in any one of paragraphs A, C, and D, further comprising: determining a lane group including the one or more first lanes based at least on the one or more first merging points and the lateral distance between the one or more first lanes and the one or more second lanes; and determining that a first geometry associated with the lane group corresponds to a second geometry associated with the one or more first road segments, wherein the association of the one or more first lanes with the one or more first road segments is based at least on the first geometry corresponding to the second geometry.

[0228] E. The method as described in any one of paragraphs A, D, and E, further comprising: determining that the one or more first merging points correspond to the one or more second merging points based at least on one or more proximitys between the one or more first locations and the one or more second locations being less than a threshold; wherein associating the one or more first lanes with the one or more first road segments and associating the one or more second lanes with the one or more second road segments is based at least on determining that the one or more first merging points correspond to the one or more second merging points.

[0229] F. The method as described in any one of paragraphs A, E, and F, wherein the one or more operations associated with the machine include one or more of the following: determining a predicted path for the machine; determining the position of the machine with respect to the one or more first merging points; planning at least one of the paths or trajectories to be followed by the machine; or calculating one or more curvatures associated with at least one of the one or more first lanes or the one or more second lanes.

[0230] G. A system comprising: one or more processors, the one or more processors being configured to: determine one or more locations in an environment corresponding to one or more junctions associated with one or more first sensing paths and one or more second sensing paths; evaluate the one or more locations based at least on map data representing the environment, determining that the one or more first sensing paths correspond to at least one map path; and perform one or more operations associated with a machine in the environment, at least based on the association between the one or more first sensing paths and the at least one map path.

[0231] H. As described in paragraph G, the one or more processors are further configured to: perform lateral classification of the one or more first sensing paths and the one or more second sensing paths, wherein determining the one or more locations corresponding to the one or more convergence points is also based at least on the lateral classification.

[0232] I. The system as described in any one of paragraphs GH, wherein one or more first sensing paths deviate from one or more second sensing paths at one or more locations near the environment corresponding to one or more convergence points.

[0233] J. The system as described in any one of paragraphs GI, wherein the one or more processors are further configured to: determine the location of the machine with respect to the one or more locations corresponding to the one or more convergence points, wherein determining the correspondence between the one or more first sensing paths and the at least one map path is also based at least on the location of the machine.

[0234] K. The system as described in any one of paragraphs GJ, wherein: the one or more first sensing paths and the one or more second sensing paths correspond to sensing lanes in the environment detected using sensing data, and the at least one map path corresponds to a segment of a driving surface in the environment depicted in the map.

[0235] L. The system as described in any one of paragraphs GK, wherein the one or more processors are further configured to: obtain sensing data indicating a plurality of sensing paths in the environment; generate an updated version of the sensing data from which a subset of the plurality of sensing paths has been removed; and use the updated version of the sensing data to determine the one or more locations corresponding to the one or more junctions.

[0236] M. The system as described in any one of paragraphs GL, wherein the subset of the plurality of sensing paths removed from the sensing data comprises at least: one or more shoulder lanes; one or more oncoming lanes; one or more bicycle lanes; or one or more portions of the one or more first sensing paths or the one or more second sensing paths having a variance that meets or exceeds a threshold.

[0237] N. The system as described in any one of paragraphs GM, wherein the one or more processors are further configured to: determine a deviation between the one or more first sensing paths and the one or more second sensing paths, at least based on the lateral distance between the one or more first sensing paths and the one or more second sensing paths satisfying or exceeding a threshold, wherein determining the one or more locations corresponding to the one or more convergence points is at least based on determining the deviation.

[0238] O. The system as described in any one of paragraphs GN, wherein the one or more processors are further configured to: determine, at least based on the map data, a second location corresponding to the second convergence point, at which the at least one map path deviates from one or more other map paths; and determine, at least based on the distance between a first location corresponding to a first convergence point among the one or more convergence points and a second location corresponding to the second convergence point being less than a threshold, a first convergence point corresponding to the second convergence point, wherein determining that the one or more first sensing paths correspond to the at least one map path is also based at least on determining that the first convergence point corresponds to the second convergence point.

[0239] P. As described in any one of paragraphs GO, the one or more processors are further configured to: determine that a first geometry associated with the one or more first sensing paths corresponds to a second geometry associated with the at least one map path, wherein determining that the one or more first sensing paths correspond to the at least one map path is also based at least on determining that the first geometry corresponds to the second geometry.

[0240] Q. A system according to any one of paragraphs GP, wherein the system comprises at least one of: a control system for an autonomous or semi-autonomous machine; a perception system for an autonomous or semi-autonomous machine; a system for performing one or more simulated operations; a system for performing one or more digital twin operations; a system for performing optical transmission simulation; a system for performing collaborative content creation of 3D assets; a system for performing one or more deep learning operations; a system implemented using an edge device; a system implemented using a robot; a system for performing one or more generative AI operations; a system for performing operations using a large language model; a system for performing operations using one or more visual language models (VLMs); a system for performing operations using one or more multimodal language models; a system for implementing one or more machine learning models as inference microservices using one or more operating system (OS) level virtualization packages; a system for performing one or more conversational AI operations; a system for generating synthetic data; a system for presenting at least one of virtual reality content, augmented reality content, or mixed reality content; a system comprising one or more virtual machines (VMs); a system implemented at least partially in a data center; or a system implemented at least partially using cloud computing resources.

[0241] R. One or more processors, comprising: processing circuitry for evaluating one or more path matching algorithms in a simulation rendered using one or more optical transmission simulation algorithms, the one or more path matching algorithms being configured to determine one or more junction points associated with the plurality of sensing paths using one or more lateral distances between the plurality of sensing paths, and to use the one or more junction points to match at least one sensing path among the plurality of sensing paths with at least one map path.

[0242] S. One or more processors as described in paragraph R, wherein the simulation is generated at least in part using a 3D content collaboration platform with 3D assets.

[0243] T. One or more processors as described in any one of paragraphs RS, wherein the 3D content collaboration platform for the 3D assets uses generic scene descriptor (USD) data to manage one or more attributes of the simulation environment associated with the simulation.

Claims

1. A method comprising: Using perception data and based at least on the lateral distance between one or more first lanes and one or more second lanes, determine one or more first locations in the environment corresponding to one or more first merging points, at which the one or more first lanes deviate from the one or more second lanes; Based at least on map data representing the environment, determine one or more second locations in the environment corresponding to one or more second junctions, at which one or more first road segments deviate from one or more second road segments; At least one or more first locations and one or more second locations are used to associate one or more first lanes with one or more first road segments and to associate one or more second lanes with one or more second road segments; as well as At least based on the association, perform one or more operations associated with the machine in the environment.

2. The method of claim 1, further comprising: The sensed data is preprocessed to remove one or more features from the sensed data, the one or more features including at least one of the following: One or more shoulder lanes; One or more oncoming lanes; One or more bike lanes; or The one or more first lanes or one or more portions of the one or more second lanes having a variance that meets or exceeds a threshold.

3. The method of claim 1, wherein determining the one or more first locations in the environment corresponding to the one or more first meeting points comprises: Based at least on the perception data, it is detected that the second lateral distance between the one or more first lanes and the one or more second lanes meets or exceeds a first threshold after the one or more first merging points; as well as The one or more first locations corresponding to the one or more first meeting points are determined based at least on the fact that the lateral distance is less than a second threshold near the one or more first locations.

4. The method of claim 1, further comprising: A lane group including the one or more first lanes is determined based at least on the one or more first merging points and the lateral distances between the one or more first lanes and the one or more second lanes; and The first geometry associated with the lane group is determined to correspond to a second geometry associated with the one or more first road segments. The association of the one or more first lanes with the one or more first road segments is at least based on the first geometry corresponding to the second geometry.

5. The method of claim 1, further comprising: The one or more first meeting points correspond to the one or more second meeting points, based at least on one or more proximitys between the one or more first locations and the one or more second locations being less than a threshold. The association of one or more first lanes with one or more first road segments and the association of one or more second lanes with one or more second road segments are based at least on determining that one or more first merging points correspond to one or more second merging points.

6. The method of claim 1, wherein the one or more operations associated with the machine include one or more of the following: Determine the predicted path for the machine; Determine the position of the machine with respect to the one or more first meeting points; Plan at least one of the paths or trajectories that the machine should follow; or Calculate one or more curvatures associated with at least one of the one or more first lanes or the one or more second lanes.

7. A system comprising: One or more processors, said one or more processors being used for: Identify one or more locations in the environment that correspond to one or more convergence points associated with one or more first sensing paths and one or more second sensing paths; The one or more locations are evaluated based at least on map data representing the environment, and the one or more first perception paths correspond to at least one map path; as well as At least based on the association of the one or more first perception paths with the at least one map path, perform one or more operations associated with the machine in the environment.

8. The system of claim 7, wherein the one or more processors are further configured to: perform lateral classification on the one or more first sensing paths and the one or more second sensing paths, wherein determining the one or more locations corresponding to the one or more convergence points is also based at least on the lateral classification.

9. The system of claim 7, wherein one or more first sensing paths deviate from one or more second sensing paths at one or more locations in the environment corresponding to one or more convergence points.

10. The system of claim 7, wherein the one or more processors are further configured to: determine the location of the machine with respect to the one or more locations corresponding to the one or more rendezvous points, wherein determining the correspondence between the one or more first sensing paths and the at least one map path is also based at least on the location of the machine.

11. The system of claim 7, wherein: The one or more first sensing paths and the one or more second sensing paths correspond to sensing lanes in the environment detected using sensing data, and The at least one map path corresponds to a segment of the driving surface in the environment depicted in the map.

12. The system of claim 7, wherein the one or more processors are further configured to: Obtain perception data indicating multiple perception paths in the environment; Generate an updated version of the sensing data from which a subset of the plurality of sensing paths has been removed; and The updated version of the sensed data is used to determine the one or more locations corresponding to the one or more convergence points.

13. The system of claim 12, wherein the subset of the plurality of sensing paths removed from the sensing data comprises at least: One or more shoulder lanes; One or more oncoming lanes; One or more bicycle lanes; or The one or more first sensing paths or one or more portions of the one or more second sensing paths having a variance that meets or exceeds a threshold.

14. The system of claim 7, wherein the one or more processors are further configured to: The deviation between the one or more first sensing paths and the one or more second sensing paths is determined based on at least the lateral distance between the one or more first sensing paths and the one or more second sensing paths satisfying or exceeding a threshold. The determination of the one or more positions corresponding to the one or more meeting points is based at least on the determination of the deviation.

15. The system of claim 7, wherein the one or more processors are further configured to: Based at least on the map data, at least one second location corresponding to the second rendezvous point is determined, at which the at least one map path deviates from one or more other map paths; and The first merging point corresponding to the second merging point is determined based at least on the fact that the distance between the first position corresponding to the first merging point among the one or more merging points and the second position corresponding to the second merging point is less than a threshold. The determination that one or more first sensing paths correspond to the at least one map path is also based at least on the determination that the first meeting point corresponds to the second meeting point.

16. The system of claim 7, wherein the one or more processors are further configured to: The first geometry associated with the one or more first perceived paths is determined to correspond to a second geometry associated with the at least one map path. The determination that the one or more first perception paths correspond to the at least one map path is also based at least on determining that the first geometry corresponds to the second geometry.

17. The system of claim 7, wherein the system comprises at least one of the following: Control systems for autonomous or semi-autonomous machines; Sensing systems for autonomous or semi-autonomous machines; A system for performing one or more simulation operations; A system for performing one or more digital twin operations; A system for performing optical transmission simulation; A system for collaborative content creation of 3D assets; A system for performing one or more deep learning operations; Systems implemented using edge devices; Systems implemented using robots; A system for performing one or more generative AI operations; A system for performing operations using a large language model; A system for performing operations using one or more visual language models (VLMs); A system for performing operations using one or more multimodal language models; A system that uses one or more operating system OS-level virtualization packages to implement one or more machine learning models as inference microservices; A system for performing one or more conversational AI operations; A system for generating synthetic data; A system for presenting at least one of virtual reality content, augmented reality content, or mixed reality content; A system containing one or more virtual machines (VMs); A system that is at least partially implemented in a data center; or A system that utilizes cloud computing resources at least in part.

18. One or more processors, comprising: The processing circuitry is configured to evaluate one or more path matching algorithms in a simulation rendered using one or more optical transmission simulation algorithms, the one or more path matching algorithms being configured to determine one or more junction points associated with the plurality of sensing paths using one or more lateral distances between the plurality of sensing paths, and to use the one or more junction points to match at least one sensing path among the plurality of sensing paths with at least one map path.

19. One or more processors as claimed in claim 18, wherein the simulation is generated at least in part using a 3D content collaboration platform for 3D assets.

20. The processor of claim 19, wherein the 3D content collaboration platform for the 3D assets uses Universal Scene Descriptor (USD) data to manage one or more attributes of the simulation environment associated with the simulation.

Images