Dynamic object detection using lidar data in autonomous machine systems and applications

Abstract

Systems are disclosed that track objects in an environment using projection images generated from a LiDAR. A deep neural network (DNN) computes a motion mask that indicates motion corresponding to points representing objects in the environment. The environment (200) includes a location of the ego machine at T-1 (202A), a location of the ego machine at TO (202B), an ego trajectory (204), a vehicle location (206A), a vehicle location (206B), a wall (208), and 3D points (210), (212), and (214). If the system determines that a depth value corresponding to a 3D point, such as 3D point (210), has changed over time, the system can infer that motion has occurred at that particular 3D point. If a first measured distance for a pixel from the location of the ego machine at T-1 (202A) to 3D point (214) on the wall (208) is 10 meters in a previous range image at time T-1, while a second measured distance for that pixel from the location of the ego machine at TO (202B) to 3D point (210) is 5 meters in a current range image at time TO, the system can determine that the vehicle has moved from vehicle location (206A) to vehicle location (206B) and is now blocking the line of sight of one or more sensors of the ego machine at location (202B) such that the ego machine is unable to measure a distance from location (202B) to 3D point (212). The projections can be based on tracked ego motion.

Description

Background Technology

[0001] The ability to detect static and dynamic features and objects in a safe environment is a crucial task for any autonomous or semi-autonomous system, such as autonomous or semi-autonomous driving systems. For example, by observing motion (or the absence of motion) across frames, static features or dynamic actors can be identified to assist in various downstream tasks, such as object tracking, path planning, obstacle avoidance, control decisions, and / or similar tasks.

[0002] Some conventional methods use object detectors, which can be executed across one or more data frames to match objects or features across frames, and then infer motion from the matched features or objects. However, such conventional methods are limited by the object detector's ability to identify objects, which requires prior knowledge of each type of object or feature to be detected. For example, these systems typically require extensive training using large amounts of training data that include descriptions of the specific objects the object detector is trained to detect. However, given the existence of a virtually infinite number of object types, these conventional methods are often insufficient and / or incomplete solutions for detecting and tracking objects.

[0003] Other systems may employ optical flow methods to find pixel-level flow fields from one frame to subsequent frames based on the time series of analyzed frames. These conventional methods require each frame (e.g., each LiDAR range image) to include sufficient texture information to allow for accurate tracking across frames. However, generating data with sufficient texture information (especially LiDAR data) is challenging due to LiDAR sensor viewpoint variations and / or potential scene occlusion. While some conventional systems can combine the above methods, these combinations do not overcome many of the drawbacks of these conventional solutions. Summary of the Invention

[0004] Embodiments of this disclosure relate to detecting static and dynamic features from LiDAR in autonomous machine applications. Systems and methods for determining motion based on one or more range images and one or more projected images are disclosed.

[0005] Compared to conventional systems such as those described above, the systems and methods of this disclosure use point clouds and / or other projected images of detection representations (e.g., LiDAR distance images, top-down or bird's-eye view projections, etc.) to detect and / or track objects in the environment (e.g., stationary and / or moving objects). For example, machine learning models (such as deep neural networks (DNNs)) can be used to compute motion masks or other output types indicating motion corresponding to points or pixels representing objects or features in the environment. Various input channels can be provided as inputs to the machine learning model to assist in computing the output. For example, one or more projected images can be generated based on projecting depth values ​​from the current distance image to the coordinate space of a previous distance image and / or projecting depth values ​​from a previous distance image to the coordinate space of the current distance image. In some embodiments, one or more comparison images can be generated based on comparing depth values ​​projected from the current distance image to the coordinate space of the previous distance image with the depth values ​​of the previous distance image. When performing a projection from one coordinate space to another, the projection can be based on the tracked ego motion (e.g., the recorded ego machine's motion between a time associated with a previous frame and a time associated with the current frame). Furthermore, the current distance image and / or the previous distance image can be directly provided as input to the machine learning model. Therefore, the machine learning model can use one or more projected images, one or more comparison images, and / or one or more distance images (or other input representations, such as top-view projected images) to compute motion masks and / or motion vector output representations.

[0006] Machine learning models can be lightweight due to the organization and quality of information in their input channels. For example, when the machine learning model is a convolutional neural network (CNN), the CNN may only require very local convolutional support—for example, it may only need ten or fewer layers (e.g., six layers in total in some embodiments). Furthermore, a CNN may include only convolutional layers, excluding, for example, fully connected layers or other layer types that might require more computation. To provide additional support for frames where occlusion, shadows, and / or noise may attenuate the input signal, input channels can be computed between two or more frames (e.g., between two or more previous and current frames). Thus, by including additional input channels corresponding to multiple previous frames, the machine learning model can obtain additional data when processing accurate or precise outputs that take into account noise, occlusion, and / or shadows. Attached Figure Description

[0007] The following describes in detail, with reference to the accompanying drawings, a system and method for detecting static and dynamic features from LiDAR in autonomous machine applications, wherein:

[0008] Figure 1This is a data flow diagram of a static and dynamic feature detection system according to some embodiments of the present disclosure;

[0009] Figure 2 Example driving environments according to some embodiments of this disclosure are described;

[0010] Figure 3 Example input images according to some embodiments of this disclosure are described;

[0011] Figure 4 This is a flowchart illustrating a method for calculating data indicating motion at a distance of one or more pixels from an image, according to some embodiments of the present disclosure;

[0012] Figure 5A These are illustrations of example autonomous vehicles according to some embodiments of the present disclosure;

[0013] Figure 5B According to some embodiments of this disclosure Figure 5A Examples of camera positions and fields of view for autonomous vehicles;

[0014] Figure 5C According to some embodiments of this disclosure Figure 5A A block diagram of an example system architecture for an example autonomous vehicle;

[0015] Figure 5D Cloud-based servers and according to some embodiments of this disclosure Figure 5A A system diagram illustrating communication between autonomous vehicles;

[0016] Figure 6 This is a block diagram of an example computing device suitable for implementing some embodiments of the present disclosure; and

[0017] Figure 7 This is a block diagram of an example data center applicable to implementing some embodiments of this disclosure. Detailed Implementation

[0018] Systems and methods relating to detecting static and dynamic features from LiDAR in autonomous machine applications are disclosed. While this disclosure describes an example autonomous vehicle 500 (also referred to herein as "vehicle 500" or "ego-vehicle 500"), its examples pertain to... Figures 5A-5D(Described herein), but this is not intended to be limiting. For example, the systems and methods described herein can be used by, but are not limited to, non-autonomous vehicles, semi-autonomous vehicles (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 vehicles, motorcycles, electric or motorized bicycles, aircraft, construction vehicles, underwater vehicles, drones, and / or other vehicle types. Furthermore, while this disclosure may describe feature and / or object tracking in autonomous machine applications, 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 surveillance, autonomous or semi-autonomous machine applications, and / or any other technological space where feature and / or object tracking can be used.

[0019] In some embodiments, as the ego machine traverses its environment, one or more of its sensors (e.g., LiDAR, RADAR, etc.) can generate a series of frames or sensor data that can be converted into one or more representations. For example, when using a LiDAR sensor, raw LiDAR data can be generated at each frame or time step, which can be used to generate a point cloud in 3D world space representing the depth, height, and lateral position of points corresponding to features and / or objects in the environment. For example, the point cloud can be used to generate one or more 2D image representations (such as LiDAR distance images, top-down or bird's-eye view images, etc.) representing the height and lateral position of points at one or more (x,y) pixel locations, where one or more depth values ​​(and / or other values, such as intensity, reflectivity, etc.) are encoded into pixels. One or more depth values ​​may correspond to distances from one or more sensors to 3D points in the environment represented by the distance images.

[0020] Distance images of the current frame and one or more previous frames can be directly used as input channels for machine learning models (e.g., deep neural networks (DNNs)) or used to generate one or more input channels for machine learning models. For example, in some embodiments, depth values ​​corresponding to the current distance image can be transformed or projected into the coordinate system of the previous distance images to generate a projected image. For example, a transformation can be applied to 3D points corresponding to one or more depth values ​​in the current distance image to locate these 3D points within the pixel indices of one or more previous distance images using the coordinate system of one or more previous distance images. As another example, one or more depth values ​​corresponding to the previous distance image can be transformed into the coordinate space of the current distance image to generate a projected image.

[0021] In some embodiments, the input channel can be generated by projecting the current distance image onto the coordinate system of the previous image, comparing the depth values ​​of one or more projections with the depth values ​​of one or more previous images, and then generating a comparison image of the changes in depth values ​​between encoded frames. Where depth values ​​are transformed or projected to another coordinate space corresponding to different frames, the transformation or projection can be performed using ego motion information (e.g., rotation, position, and / or velocity data captured by the ego machine's IMU, GPS, and / or visual odometry system). Thus, once the depth values ​​of a particular pixel differ (e.g., exceed a threshold depth difference) in the same coordinate system, this can indicate that the point or pixel corresponds to a moving or dynamic object. In some embodiments, the depth values ​​and positions (e.g., order) of 3D points in the current distance image can be saved for use in subsequent calculations, where the current distance image can be used as a previous distance image employing the corresponding coordinate system. In addition to or instead of one or more projected or comparison images, the previous distance image and / or the current distance image can also be provided directly as input channels to the machine learning model without projecting to different coordinate spaces.

[0022] The number of input channels can depend on the number of previous frames corresponding to the previous timestamps provided to the machine learning model. For example, the number of channels provided to the machine learning model in the input image can be calculated according to equation (1), as follows:

[0023] NUM_CHANNELS = 3 * NUM_FRAMES + 1 (1)

[0024] Where NUM_FRAMES is the number of previous frames at the previous timestamp. For each previous frame, the three input channels may include the current frame projected onto the previous frame, the previous frame projected onto the current frame, and a comparison image comparing the value projected onto the current frame with the value of the previous frame. Although described as including all three channels for each frame, this is not intended to be limiting, and in some embodiments, one or more of these channels may be available for each previous frame.

[0025] Regarding a single 3D point in the environment, a channel can represent a first channel associated with the distance between that 3D point and the distance between the self-machine at position T0 and the 3D point of interest. This first channel can correspond to directly applying the distance image corresponding to the current time as the input channel and can correspond to "+1" in equation (1) above. A second channel can be associated with the distance between the self-machine at position T0 and a first 3D point at T-1, which, when viewed from the self-machine at position T0, projects to the same pixel as the 3D point of interest. A third channel can be associated with the distance between the self-machine at position T-1 and the 3D point of interest. A fourth channel can be associated with the distance between the self-machine at position T-1 and a second 3D point at T-1, which, when viewed from the self-machine at position T-1, projects to the same pixel as the 3D point of interest.

[0026] In some embodiments, channels may be provided as one or more input images (e.g., a channel stack) to a machine learning model, which may process the channel stack to generate a motion mask having motion confidence values ​​(e.g., from 0 to 1) assigned to one or more (e.g., each) of a plurality of pixels, indicating the confidence of a pixel in being associated with a static or dynamic object or feature; for example, a pixel with a value of 0.3 may be less likely to have associated motion than a pixel with a value of 0.9. In some embodiments, using the motion mask output from the machine learning model, the system may identify clusters of points corresponding to objects and track these objects across frames. In a further embodiment, the motion mask may be used to identify regions of interest and may be provided as input to an independent object detector, which may increase the detection rate of the independent object detector. Furthermore, in addition to motion confidence values, the system may further determine vectors corresponding to 3D points by tracking pixel motion between frames. In some embodiments, in addition to or in lieu of confidence values, the machine learning model may also be trained to compute output motion vectors. For example, for a given pixel in the current frame, the output can include a motion vector pointing to the same pixel in the previous frame (e.g., a pixel corresponding to the same feature or object).

[0027] When the machine learning model is a DNN (such as a convolutional neural network (CNN)), the DNN can include a lightweight architecture, such as a fully convolutional architecture consisting only of convolutional layers. In some embodiments, the DNN may have ten or fewer layers in total (e.g., six layers). Due to the amount of information and detail available in the input channels according to embodiments of this disclosure, a limited number of layers and an overall lightweight structure are possible.

[0028] Therefore, if the system determines that the depth value corresponding to a 3D point has changed over time, the system can infer that motion has occurred at that particular 3D point. As a non-limiting example, if the distance from the ego machine to a point on the wall is 10 meters in a previous distance image at time T-1, and the distance from the ego machine to that point on the wall is 5 meters in the current distance image at time T0, then the system can determine that the object has moved into the line of sight of one or more sensors of the ego machine.

[0029] refer to Figure 1 , Figure 1 This is an example data flow diagram for system 100 to detect static and dynamic features from sensor data in an autonomous machine application, according to some embodiments of this disclosure. It should be understood that such and other arrangements described herein are illustrative only. Other arrangements and elements (e.g., machines, interfaces, functions, sequences, functional groups, etc.) may be used in addition to or instead of those 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 implemented in any suitable combination and location. The various functions described herein as being performed by entities can be performed by any hardware, firmware, and / or software (or combinations thereof). For example, various functions can be performed by a processor executing instructions stored in memory. In some embodiments, system 100 may use... Figures 5A-5D Example of autonomous vehicles 500 Figure 6 Example computing device 600, and / or Figure 7 The example data center 700 uses similar components, features, and / or functions to achieve this.

[0030] Figure 1The data stream includes sensor data 110, channel generator 120, point cloud generator 122, image comparator 124, one or more channels 130, motion model 140, and motion mask 150. In some embodiments, sensor data 110 may include, but is not limited to, sensor data 110 from any sensor of vehicle 500 (and / or other vehicles, machines, or objects, such as robotic equipment, ships, aircraft, trains, construction equipment, VR systems, AR systems, etc., as some examples). For non-limiting examples, such as where one or more sensors generating sensor data 110 are mounted on or otherwise associated with the vehicle, sensor data 110 may include data generated by, but not limited to, one or more Global Navigation Satellite System (GNSS) sensors 558 (e.g., one or more Global Positioning System sensors), one or more RADAR sensors 560, one or more ultrasonic sensors 562, one or more LIDAR sensors 564, and one or more inertial measurement unit (IMU) sensors. 566 (e.g., one or more accelerometers, one or more gyroscopes, one or more magnetic compasses, one or more magnetometers, etc.), one or more microphones 596, one or more stereo cameras 568, one or more wide-angle cameras 570 (e.g., fisheye cameras), one or more infrared cameras 572, one or more surround cameras 574 (e.g., 360-degree cameras), one or more long-range and / or mid-range cameras 598, one or more speed sensors 544 (e.g., for measuring the speed of vehicle 500), and / or other sensor types.

[0031] In some embodiments, as the vehicle 500 travels through its environment, one or more sensors of the vehicle 500 (e.g., LiDAR sensors, RADAR sensors, etc.) can generate sensor data 110. Sensor data 110 can be passed to channel generator 120 and converted into one or more representations. For example, as the vehicle 500 travels through its environment, sensor data 110 (such as raw LiDAR data) can be generated at each frame or time step. Point cloud generator 122 can then use the sensor data 110 from each frame or time step to generate a point cloud in 3D world space, representing the depth, height, and lateral position of points corresponding to features and / or objects in the environment. For example, using the generated point cloud, channel generator 120 can generate one or more 2D image representations (e.g., LiDAR distance images) that can represent the height and lateral position of points at (x,y) pixel locations, where depth values ​​(and / or other values, such as intensity, reflectivity, etc.) are encoded as pixels. The depth value represented by the image can correspond to the distance from one or more sensors of the vehicle 500 to a 3D point in the environment represented by the distance image.

[0032] In some embodiments, distance images corresponding to the current frame (or time) and one or more previous frames (or times) may be output as one or more channels 130. These channels 130 may be provided to a motion model 140, such as a deep neural network (DNN). Alternatively or additionally, the distance images may be manipulated by a channel generator 120 to generate one or more channels 130 for the motion model 140. For example, in some embodiments, the channel generator 120 may transform or project one or more depth values ​​corresponding to the current distance image into the coordinate system of the previous distance images to generate a projected image. For example, the channel generator 120 may apply a transformation to 3D points corresponding to one or more depth values ​​in the current distance image to locate these 3D points within the pixel indices of one or more previous distance images using the coordinate system of one or more previous distance images. In a further embodiment, the channel generator 120 may transform depth values ​​corresponding to previous distance images into the coordinate space of the current distance image to generate a projected image.

[0033] In some embodiments, image comparator 124 generates channel 130 by projecting the current distance image onto the coordinate system of a previous image, comparing the projected depth values ​​with the depth values ​​of the previous image, and then generating a comparison image of the changes in depth values ​​between encoded frames. Wherein the depth values ​​are transformed or projected to another coordinate space corresponding to different frames, sensor data 110 may include self-motion information, such as rotation, position, and / or velocity data captured by the vehicle 500's IMU, GPS, and / or visual odometry system. Using the self-motion information included in sensor data 110, channel generator 120 can transform or project the depth values ​​to another coordinate space. Thus, once in the same coordinate system, if the depth values ​​for a particular pixel differ (e.g., exceeding a threshold depth difference), image comparator 124 can determine that the point or pixel corresponds to a moving or dynamic object. In some embodiments, channel generator 120 may save the depth values ​​and positions (e.g., order) of 3D points in the current distance image for use in subsequent calculations, where the current distance image can be used as a previous distance image in the corresponding coordinate system. In addition to or as a substitute for one or more projected or comparative images, the front distance image and / or the current distance image may be provided directly to the motion model 140 as one or more channels 130 without being projected to different coordinate spaces.

[0034] In some embodiments, the number of one or more channels 130 output by channel generator 120 may depend on the number of previous frames corresponding to the previous timestamps to be provided to motion model 140. As a non-limiting example, for each previous frame, the three input channels may include a current frame projected onto the preceding frame, a previous frame projected onto the current frame, and a comparison image from image comparator 124 comparing the current frame value projected onto the preceding frame with the previous frame value. Although described as including three channels per frame, this is not intended to be limiting; in some embodiments, one or more channels may be available for each preceding frame.

[0035] In some embodiments, one or more channels 130 may be provided to a motion model 140 as one or more input images (e.g., a channel stack). The motion model 140 may process the channel stack to generate one or more motion masks 150, which have motion confidence values ​​(e.g., from 0 to 1) assigned to one or more (e.g., each) of a plurality of pixels, indicating the confidence that the pixel is associated with a static or dynamic object or feature. In some embodiments, using one or more motion masks 150 output from the motion model 140, the system 100 may identify clusters of points corresponding to objects in one or more motion masks 150 and track those objects across frames. In a further embodiment, one or more motion masks 150 may be used to identify regions of interest and may be provided as input to an independent object detector, which may increase the detection rate of the independent object detector. Furthermore, in addition to motion confidence values, the system may further determine vectors corresponding to 3D points by tracking pixel motion between frames. In some embodiments, in addition to or in lieu of confidence values, the motion model 140 may also be trained to compute output motion vectors. For example, for a given pixel in the current frame, the output can include a motion vector pointing to the same pixel in the previous frame (e.g., a pixel corresponding to the same feature or object).

[0036] Now for reference Figure 2 , Figure 2 This is an example environment 200 according to some embodiments of the present disclosure. Environment 200 includes a self-machine at position 202A of T-1, a self-machine at position 202B of T0, a self-track 204, vehicle positions 206A and 206B, a wall 208, and 3D points 210, 212, and 214. Environment 200 shows the self-machine traveling along self-track 204 from position 202A of T-1 to position 202B of T0. Furthermore, Figure 2 The effect of measuring the distance to the object for the pixel is shown when the self-machine moves from position 202A of T-1 to position 202B of T0, and the vehicle moves from position 206A of vehicle to position 206B of vehicle.

[0037] Regarding 3D point 210 in environment 200, the generated channels can represent a first channel associated with the distance between 3D point 210 and the self-machine's position 202B at T0. This first channel can correspond to directly applying a distance image corresponding to the current time as an input channel. A second channel can be associated with the distance between the self-machine's position 202B at T0 and 3D point 212 at T-1, which, when viewed from the self-machine's position 202B at T0, projects to the same pixel as 3D point 210. A third channel can be associated with the distance between the self-machine's position 202A at T-1 and 3D point 210. A fourth channel can be associated with the distance between the self-machine's position 202A at T-1 and 3D point 214 at T-1, which, when viewed from the self-machine's position 202A at T-1, projects to the same pixel as 3D point 210.

[0038] Therefore, if the system determines that the depth value corresponding to a 3D point (such as 3D point 210) has changed over time, the system can infer that motion has occurred at that particular 3D point. As a non-limiting example, if the first measured distance for a pixel from the self-machine at position 202A at time T-1 to an object (e.g., 3D point 214 on wall 208) is 10 meters in a previous distance image at time T-1, and the second measured distance for that pixel from the self-machine at position 202B at time T0 to the object (e.g., 3D point 210) is 5 meters in the current distance image at time T0, then the system can determine that the vehicle has moved from vehicle position 206A to vehicle position 206B and is now blocking the line of sight of one or more of the self-machine's sensors at position 202B at time T0. In other words, at T0, because the vehicle has moved from vehicle position 206A to vehicle position 206B, the self-machine cannot measure the distance from position 202B to 3D point 212. Conversely, the distance measured for this pixel is the distance from position 202B to 3D point 210, and based on this distance difference, the system can determine the motion associated with this pixel.

[0039] Now for reference Figure 3 , Figure 3 This is an example input image 300 according to some embodiments of the present disclosure. Input image 300 includes channels 302, 304, 306, and 308. Although Figure 3Four channels are shown, but the number of channels provided to the machine learning model (e.g., motion model 140) in the input image 300 can be calculated according to equation (1) above. In some examples, each preceding frame may include three input channels, which may include a current frame projected onto the preceding frame, a preceding frame projected onto the current frame, and a comparison image comparing the current frame value projected onto the preceding frame with the preceding frame value. Although described as including all three channels for each preceding frame, this is not intended to be limiting, and in some embodiments, one or more of these channels may be available for each preceding frame.

[0040] To use Figure 2 Taking environment 200, locations 202A / B, and 3D points 210, 212, and 214 as examples, each of channels 302, 304, 306, and 308 can include distance information. For instance, channel 302 can include the distance between the self-machine's location 202B at T0 and 3D point 210; channel 304 can include the distance between the self-machine's location 202B at T0 and 3D point 212; channel 306 can include the distance between the self-machine's location 202A at T-1 and 3D point 210; and channel 308 can include the distance between the self-machine's location 202A at T-1 and 3D point 214. An input image 300 with each of these channels and corresponding distance information can be provided to a machine learning model to determine static and dynamic objects in the environment, such as... Figure 2 The environment in 200.

[0041] When the machine learning model is a DNN, such as a convolutional neural network (CNN), the DNN can include a lightweight architecture, such as a fully convolutional architecture consisting only of convolutional layers. In some embodiments, the DNN may have ten or fewer layers in total (e.g., six layers). Due to the amount of information and detail available in the input image 300, a limited number of layers and an overall lightweight architecture are possible.

[0042] While this article describes examples of using DNNs, particularly Convolutional Neural Networks (CNNs), it is not intended to be limiting. For example, but not limited to, DNN 126 can include one or more machine learning models of any type, such as those using linear regression, logistic regression, decision trees, support vector machines (SVMs), Naive Bayes, k-nearest neighbors (Knn), K-means clustering, random forests, dimensionality reduction algorithms, gradient boosting algorithms, neural networks (e.g., autoencoders, convolutions, recurrent neural networks, perceptrons, long / short-term memory / LSTMs, Hopfield, Boltzmann, deep belief, deconvolution, generative adversarial, liquid state machines, etc.), region of interest detection algorithms, computer vision algorithms, and / or other types of machine learning models.

[0043] In some embodiments, when using a DNN, the DNN may include any number of layers. One or more layers may include convolutional layers. Convolutional layers can compute the outputs of neurons connected to local regions in the input layer, with each neuron computing the dot product between its weights and the small region in the input volume it is connected to. One or more of these layers may include rectified linear unit (ReLU) layers. One or more ReLU layers may apply element-wise activation functions, such as max(0,x), for example, with a threshold set to 0. The resulting volume of the ReLU layer may be the same as the volume of the input to the ReLU layer. One or more of these layers may include pooling layers. Pooling layers may perform downsampling operations along spatial dimensions (e.g., height and width), which can result in a volume smaller than the input to the pooling layer. One or more of these layers may include one or more fully connected layers. Each neuron in one or more fully connected layers may be connected to each neuron in the previous volume. The fully connected layers may compute class scores, and the resulting volume may be 1 × 1 × number of classes. In some examples, a CNN may include one or more fully connected layers, such that the outputs of one or more layers of the CNN can be fed as input to one or more fully connected layers of the CNN. In some examples, a DNN may implement one or more convolutional streams, some or all of which may include one or more fully connected layers, each of which may be a fully connected layer. In some non-limiting embodiments, a DNN may include a series of convolutional layers and max-pooling layers for image feature extraction, followed by multi-scale dilated convolutional layers and upsampling layers to facilitate global contextual feature extraction.

[0044] While this paper discusses input layers, convolutional layers, pooling layers, ReLU layers, and fully connected layers in relation to DNNs, this is not intended to be limiting. For example, additional or alternative layers, such as normalization layers, SoftMax layers, and / or other layer types, can be used in DNNs. In embodiments where DNNs include CNNs, different orders and numbers of CNN layers may be used depending on the embodiment. In other words, the order and number of layers in a DNN are not limited to any single architecture.

[0045] Furthermore, some layers may include parameters (such as weights and / or biases), such as convolutional and fully connected layers, while other layers may not include parameters, such as ReLU layers and pooling layers. In some examples, the parameters can be learned by the DNN during training. Additionally, some layers may include additional hyperparameters (e.g., learning rate, stride, epochs, etc.), such as convolutional, fully connected, and pooling layers, while other layers may not include additional hyperparameters, such as ReLU layers. Parameters and hyperparameters are not limited and may vary depending on the implementation.

[0046] Now for reference Figure 4 Each block of the method 400 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. The method can also be embodied as computer-usable instructions stored on a computer storage medium. The method can be provided by a standalone application, service, or managed service (alone or in combination with another managed service) or a plug-in to another product, to name a few. Furthermore, by way of example, regarding... Figure 1 The system for detecting static and dynamic features describes method 400. However, these methods may be additionally or alternatively performed by any single system or any combination of systems, including but not limited to those described herein.

[0047] Figure 4 This is a flowchart illustrating a method 400 for calculating data indicating motion at one or more pixels in a distance image, according to some embodiments of the present disclosure. At block B402, method 400 includes: generating a first distance image at a first time step and a second distance image at a subsequent second time step using a LiDAR sensor of a self-operated machine. For example, when using a LiDAR sensor, raw LiDAR data can be generated at each frame or each time step, which can be used to generate a point cloud in 3D world space representing the depth, height, and lateral position of points corresponding to features and / or objects in the environment. For example, using the point cloud, one or more 2D image representations (such as LiDAR distance images, top-view or bird's-eye view images, etc.) can be generated representing the height and lateral position of a point at a (x,y) pixel location, where depth values ​​(and / or other values, such as intensity, reflectivity, etc.) are encoded as pixels.

[0048] At box B404, method 400 includes: generating a first projected image based at least in part on projecting a first depth value from a first distance image onto a first coordinate space of a second distance image; and generating a second projected image based at least in part on projecting a second depth value from the second distance image onto a second coordinate space of the first distance image. For example, one or more depth values ​​corresponding to the current distance image can be transformed or projected onto the coordinate system of a previous distance image to generate the projected image. For example, a transformation can be applied to 3D points corresponding to depth values ​​in the current distance image to locate these 3D points within pixel indices of one or more previous distance images using the coordinate systems of one or more previous distance images.

[0049] At box B406, method 400 includes: using a deep neural network (DNN) to compute data indicating motion at one or more pixels in the second distance image, based at least in part on the second distance image, the first projection image, and the second projection image. For example, channels may be provided as one or more input images (e.g., a channel stack) to a motion model 140 (e.g., a machine learning model), which may process the channel stack to generate a motion mask having motion confidence values ​​(e.g., from 0 to 1) assigned to one or more (e.g., each) of a plurality of pixels, indicating the confidence of a pixel in being associated with a static or dynamic object or feature; for example, a pixel with a value of 0.3 may be less likely to have associated motion than a pixel with a value of 0.9.

[0050] Example autonomous vehicles

[0051] Figure 5A This is an illustration of an example autonomous vehicle 500 according to some embodiments of the present disclosure. The autonomous vehicle 500 (or, alternatively, referred to herein as “vehicle 500”) may include, but is not limited to, passenger vehicles such as automobiles, trucks, buses, ambulances, shuttles, electric or motorized bicycles, motorcycles, fire trucks, police cars, ambulances, boats, engineering vehicles, underwater vessels, drones, trailer-coupled vehicles, and / or other types of vehicles (e.g., driverless and / or capable of accommodating one or more passengers). Autonomous vehicles are generally described according to the level 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) “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 500 may be able to perform one or more functions that meet the requirements of Level 3 to Level 5 autonomous driving. For example, depending on the embodiment, vehicle 500 may be able to perform conditional automation (Level 3), high automation (Level 4), and / or full automation (Level 5).

[0052] Vehicle 500 may include components such as chassis, body, wheels (e.g., 2, 4, 6, 8, 18, etc.), tires, axles, and other vehicle components. Vehicle 500 may include a propulsion system 550, such as an internal combustion engine, a hybrid power plant, an all-electric motor, and / or another type of propulsion system. Propulsion system 550 may be connected to the drivetrain of vehicle 500, which may include a transmission, to enable propulsion of vehicle 500. Propulsion system 550 may be controlled in response to receiving a signal from throttle / accelerator 552.

[0053] A steering system 554, which may include a steering wheel, can be used to steer the vehicle 500 (e.g., along a desired path or route) when the propulsion system 550 is operating (e.g., when the vehicle is in motion). The steering system 554 may receive signals from the steering actuator 556. For fully automatic (level 5) functionality, the steering wheel may be optional.

[0054] The brake sensor system 546 can be used to operate the vehicle brakes in response to receiving signals from the brake actuator 548 and / or the brake sensor.

[0055] It may include one or more System-on-a-Chip (SoC) 504 ( Figure 5C One or more controllers 536, 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 500. For example, one or more controllers may send signals to operate vehicle brakes via one or more brake actuators 548, to operate steering system 554 via one or more steering actuators 556, and to operate propulsion system 550 via one or more throttles / accelerators 552. One or more controllers 536 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 500. One or more controllers 536 may include a first controller 536 for autonomous driving functions, a second controller 536 for functional safety functions, a third controller 536 for artificial intelligence functions (e.g., computer vision), a fourth controller 536 for infotainment functions, a fifth controller 536 for redundancy in emergency situations, and / or other controllers. In some examples, a single controller 536 can handle two or more of the functions described above, and two or more controllers 536 can handle a single function, and / or any combination thereof.

[0056] One or more controllers 536 may provide signals for controlling one or more components and / or systems of vehicle 500 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, a global navigation satellite system sensor 558 (e.g., a Global Positioning System sensor), a RADAR sensor 560, an ultrasonic sensor 562, a LIDAR sensor 564, an inertial measurement unit (IMU) sensor 566 (e.g., an accelerometer, gyroscope, magnetic compass, magnetometer, etc.), a microphone 596, a stereo camera 568, a wide-angle camera 570 (e.g., a fisheye camera), an infrared camera 572, a surround camera 574 (e.g., a 360-degree camera), a long-range and / or medium-range camera 598, a speed sensor 544 (e.g., for measuring the rate of vehicle 500), a vibration sensor 542, a steering sensor 540, a braking sensor (e.g., as part of a braking sensor system 546), and / or other sensor types.

[0057] One or more of the controllers 536 may receive inputs (e.g., represented by input data) from the instrument cluster 532 of the vehicle 500 and provide outputs (e.g., represented by output data, display data, etc.) via a human-machine interface (HMI) display 534, an auditory signaling device, a speaker, and / or via other components of the vehicle 500. These outputs may include information such as vehicle speed, rate, time, map data (e.g., [missing information]). Figure 5C Information such as the HD map 522, location data (e.g., the location of vehicle 500 on the map), direction, and the location of other vehicles (e.g., occupying a grid), as well as information about objects and their states perceived by controller 536, etc. For example, HMI display 534 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.).

[0058] Vehicle 500 further includes a network interface 524, which can communicate via one or more networks using one or more wireless antennas 526 and / or a modem. For example, network interface 524 may be able to communicate via LTE, WCDMA, UMTS, GSM, CDMA2000, etc. One or more wireless antennas 526 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 LE, Z-Wave, ZigBee, etc. and / or one or more low-power wide area networks (LPWAN) such as LoRaWAN, SigFox, etc.

[0059] Figure 5BFor use in accordance with some embodiments of this disclosure Figure 5A This is an example of the camera position and field of view of an autonomous vehicle 500. 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 500.

[0060] 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 500. 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.

[0061] 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).

[0062] One or more of the cameras can be mounted in mounting components such as custom-designed (3-D printed) parts to cut off stray light and reflections from inside the vehicle (e.g., reflections from the dashboard 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-3-D 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.

[0063] A camera with a field of view that includes the environment in front of the vehicle 500 (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 536 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.

[0064] A variety of cameras can be used in front-facing configurations, including monocular camera platforms such as CMOS (Complementary Metal-Oxide-Semiconductor) color imagers. Another example could be a wide-angle camera 570, 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 5B The middle image shows only one wide-angle camera, but any number of wide-angle cameras 570 can be present on the vehicle 500. Furthermore, remote cameras 598 (e.g., a pair of long-view stereo cameras) can be used for depth-based object detection, especially for objects for which neural networks have not yet been trained. Remote cameras 598 can also be used for object detection and classification, as well as basic object tracking.

[0065] One or more stereo cameras 568 may also be included in a front-mounted configuration. The stereo camera 568 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 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 568 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 568 may be used in addition to those described herein, or alternatively.

[0066] Cameras with a field of view including the side portion of the environment of vehicle 500 (e.g., side-view cameras) can be used for surround view, providing information for creating and updating occupancy grids and generating side-impact collision warnings. For example, surround camera 574 (e.g., ... Figure 5BThe four surround cameras 574 shown can be mounted on the vehicle 500. The surround cameras 574 can include a wide-angle camera 570, a fisheye camera, a 360-degree camera, 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 574 (e.g., left, right, and rear) and can utilize one or more other cameras (e.g., a forward-facing camera) as a fourth surround-view camera.

[0067] Cameras with a field of view that includes the environment behind vehicle 500 (e.g., rear-view cameras) 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 cameras 598, stereo cameras 568, infrared cameras 572, etc.).

[0068] Figure 5C For use in accordance with some embodiments of this disclosure Figure 5A The example autonomous vehicle 500 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.

[0069] Figure 5C Each component, feature, and system in vehicle 500 is illustrated as being connected via bus 502. Bus 502 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 500 used to assist in the control of various features and functions of vehicle 500, 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.

[0070] Although bus 502 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 502 is represented by a single line, this is not intended to be limiting. For example, any number of buses 502 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 502 may be used to perform different functions and / or may be used for redundancy. For example, a first bus 502 may be used for a collision avoidance function, and a second bus 502 may be used for drive control. In any example, each bus 502 may communicate with any component of vehicle 500, and two or more buses 502 may communicate with the same component. In some examples, each SoC 504, each controller 536, and / or each computer within the vehicle may have access to the same input data (e.g., input from sensors of vehicle 500) and may be connected to a common bus such as a CAN bus.

[0071] Vehicle 500 may include one or more controllers 536, such as those described herein. Figure 5A The controllers described herein. Controller 536 can be used for a wide variety of functions. Controller 536 can be coupled to any other different components and systems of vehicle 500 and can be used for the control of vehicle 500, artificial intelligence of vehicle 500, infotainment and / or the like for vehicle 500.

[0072] Vehicle 500 may include one or more System-on-Chip (SoC) 504s. SoC 504 may include a CPU 506, GPU 508, processor 510, cache 512, accelerator 514, data storage 516, and / or other components and features not shown. SoC 504 can be used to control vehicle 500 across a wide variety of platforms and systems. For example, one or more SoCs 504s may be combined with an HD map 522 in a system (e.g., the system of vehicle 500), the HD map being transmitted via a network interface 524 from one or more servers (e.g., [server name missing]). Figure 5D One or more servers (578) receive map refresh and / or updates.

[0073] CPU 506 may include CPU clusters or CPU complexes (alternatively referred to herein as "CCPLEX"). CPU 506 may include multiple cores and / or L2 cache. For example, in some embodiments, CPU 506 may include eight cores in a coherent multiprocessor configuration. In some embodiments, CPU 506 may include four dual-core clusters, each with a dedicated L2 cache (e.g., 2MB L2 cache). CPU 506 (e.g., CCPLEX) may be configured to support simultaneous cluster operation, such that any combination of clusters of CPU 506 can be active at any given time.

[0074] CPU 506 can implement power management capabilities including one or more of the following features: automatic clock gating of hardware blocks when idle to conserve 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 506 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.

[0075] GPU 508 may include an integrated GPU (or, alternatively, referred to herein as an "iGPU"). GPU 508 may be programmable and efficient for parallel workloads. In some examples, GPU 508 may use an enhanced tensor instruction set. GPU 508 may include one or more streaming microprocessors, each of which 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 508 may include at least eight streaming microprocessors. GPU 508 may use a computation application programming interface (API). Furthermore, GPU 508 may use one or more parallel computing platforms and / or programming models (e.g., NVIDIA's CUDA).

[0076] In automotive and embedded applications, the GPU 508 can be power-optimized for optimal performance. For example, the GPU 508 can be fabricated on FinFETs. However, this is not intended to be limiting, and the GPU 508 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, dispatch units, 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 for efficient execution of workloads. The streaming microprocessor can include independent thread scheduling capabilities to allow for finer-grained synchronization and cooperation between parallel threads. Streaming microprocessors can include a combination of L1 data cache and shared memory units to improve performance while simplifying programming.

[0077] The GPU 508 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.

[0078] The GPU 508 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 508 to directly access the CPU 506 page tables. In such examples, when the GPU 508 Memory Management Unit (MMU) experiences a miss, an address translation request can be transferred to the CPU 506. In response, the CPU 506 can look up the virtual-physical mapping for the address in its page tables and transfer the translation back to the GPU 508. Thus, unified memory technology can allow a single unified virtual address space for the memory of both the CPU 506 and the GPU 508, simplifying GPU 508 programming and porting applications to the GPU 508.

[0079] In addition, the GPU 508 may include access counters that track how frequently the GPU 508 accesses the memory of other processors. These access counters help ensure that memory pages are moved to the physical memory of the processor that accesses those pages most frequently.

[0080] SoC 504 may include any number of caches 512, including those described herein. For example, cache 512 may include an L3 cache available to both CPU 506 and GPU 508 (e.g., it is connected to both CPU 506 and GPU 508). Cache 512 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 4MB or more, but a smaller cache size may also be used.

[0081] SoC 504 may include one or more arithmetic logic units (ALUs) that can be used to perform processing of any of a variety of tasks or operations related to vehicle 500, such as processing a DNN. Additionally, SoC 504 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 104 may include one or more FPUs integrated as execution units within CPU 506 and / or GPU 508.

[0082] SoC 504 may include one or more accelerators 514 (e.g., hardware accelerators, software accelerators, or combinations thereof). For example, SoC 504 may include a hardware acceleration 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 acceleration cluster to accelerate neural networks and other computations. The hardware acceleration cluster can be used to supplement GPU 508 and offload some tasks from GPU 508 (e.g., freeing up more cycles of GPU 508 to perform other tasks). As an example, accelerator 514 can be used for targeted workloads (e.g., perceptrons, convolutional neural networks (CNNs), etc.) that are stable enough to be easily controlled for 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).

[0083] Accelerator 514 (e.g., a hardware acceleration 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.

[0084] 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.

[0085] The DLA can perform any function of the GPU 508, and by using inference accelerators, for example, a designer can make either the DLA or the GPU 508 target any function. For example, a designer can focus the CNN processing and floating-point operations on the DLA and leave other functions to the GPU 508 and / or other accelerators 514.

[0086] Accelerator 514 (e.g., a hardware acceleration cluster) 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.

[0087] 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.

[0088] DMA enables PVA components to access system memory independently of the CPU 506. 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.

[0089] 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.

[0090] Each of the vector processors may include an instruction cache and may be coupled to dedicated memory. Consequently, 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 parallelization. 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 hardware acceleration cluster, 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.

[0091] Accelerator 514 (e.g., a hardware acceleration cluster) may include an on-chip computer vision network and SRAM to provide high-bandwidth, low-latency SRAM for accelerator 514. 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.

[0092] 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.

[0093] In some examples, SoC 504 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 for other uses. In some embodiments, one or more Tree Traversal Units (TTUs) may be used to perform one or more ray tracing-related operations.

[0094] Accelerators 514 (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 efficient in object detection and integer mathematical operations.

[0095] 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.

[0096] 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.

[0097] 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 a neural network to regress the confidence value. This neural network can take at least a subset of parameters as input, such as bounding box dimensions, ground plane estimates (e.g., from another subsystem), inertial measurement unit (IMU) sensor 566 outputs related to vehicle orientation and distance, 3D position estimates of objects obtained from the neural network and / or other sensors (e.g., LiDAR sensor 564 or RADAR sensor 560), etc.

[0098] SoC 504 may include one or more data storage units 516 (e.g., memory). Data storage units 516 may be on-chip memory of SoC 504, which may store neural networks to be executed on the GPU and / or DLA. In some examples, for redundancy and security, data storage units 516 may be large enough to store multiple instances of the neural network. Data storage units 512 may include L2 or L3 cache 512. References to data storage units 516 may include references to memory associated with PVA, DLA, and / or other accelerators 514 as described herein.

[0099] SoC 504 may include one or more processors 510 (e.g., embedded processors). Processor 510 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 504 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 504 thermal and temperature sensor management, and / or SoC 504 power state management. Each temperature sensor may be implemented as a ring oscillator whose output frequency is proportional to the temperature, and SoC 504 may use the ring oscillator to detect the temperature of CPU 506, GPU 508, and / or accelerator 514. 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 504 into a lower power state and / or place vehicle 500 into a driver-safe parking mode (e.g., safely stop vehicle 500).

[0100] The processor 510 may further 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, as well as 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.

[0101] The processor 510 may further 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.

[0102] Processor 510 may further 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.

[0103] The processor 510 may further include a real-time camera engine, which may include a dedicated processor subsystem for handling real-time camera management.

[0104] The processor 510 may further 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.

[0105] Processor 510 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 570, the surround camera 574, 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.

[0106] 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.

[0107] 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 508 does not need to continuously render new surfaces, the video image compositer can be further used for user interface components. Even when the GPU 508 is powered on and activated, performing 3D rendering, the video image compositer can be used to offload the GPU 508 to improve performance and responsiveness.

[0108] SoC 504 may further 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. SoC 504 may further 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.

[0109] SoC 504 may further include a wide range of peripheral interfaces to enable communication with peripherals, audio codecs, power management and / or other devices. SoC 504 can be used to process data from cameras and sensors (e.g., LiDAR sensor 564, RADAR sensor 560, etc., which can be connected via Gigabit Multimedia Serial Link and Ethernet), data from bus 502 (e.g., vehicle 500 speed, steering wheel position, etc.), and data from GNSS sensor 558 (connected via Ethernet or CAN bus). SoC 504 may further include a dedicated high-performance, high-capacity memory controller, which may include its own DMA engine, and which can be used to free up CPU 506 from routine data management tasks.

[0110] SoC 504 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. SoC 504 can be faster, more reliable, and even more energy- and space-efficient than conventional systems. For example, when combined with CPU 506, GPU 508, and data storage 516, accelerator 514 can provide a fast and efficient platform for Level 3-5 autonomous vehicles.

[0111] 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.

[0112] In contrast to conventional systems, the techniques described in this paper, by providing CPU complexes, GPU complexes, and hardware acceleration clusters, allow multiple neural networks to be executed simultaneously and / or sequentially, and the results to be combined to achieve Level 3–5 autonomous driving capabilities. For example, a CNN executed on a DLA or dGPU (e.g., GPU 520) 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 further 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.

[0113] 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, which 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, which informs 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 508.

[0114] 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 500. 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 504 provides security against theft and / or carjacking.

[0115] In another example, the CNN used for emergency vehicle detection and identification can use data from microphone 596 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 504 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 558. 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 562, 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.

[0116] The vehicle may include a CPU 518 (e.g., a discrete CPU or dCPU) that can be coupled to the SoC 504 via a high-speed interconnect (e.g., PCIe). The CPU 518 may include, for example, an x86 processor. The CPU 518 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 504, and / or monitoring the status and health of the controller 536 and / or the infotainment SoC 530.

[0117] Vehicle 500 may include a GPU 520 (e.g., a discrete GPU or dGPU) that can be coupled to SoC 504 via a high-speed interconnect (e.g., NVIDIA's NVLINK). GPU 520 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 on inputs from sensors of vehicle 500 (e.g., sensor data).

[0118] Vehicle 500 may further include a network interface 524, which may include one or more wireless antennas 526 (e.g., one or more wireless antennas for different communication protocols, such as cellular antennas, Bluetooth antennas, etc.). Network interface 524 can be used to enable wireless connectivity via the Internet to the cloud (e.g., with server 578 and / or other network devices), with other vehicles, and / or with computing devices (e.g., a passenger's client device). 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 500 with information about vehicles approaching vehicle 500 (e.g., vehicles in front, to the side, and / or behind vehicle 500). This functionality can be part of vehicle 500's cooperative adaptive cruise control function.

[0119] Network interface 524 may include a SoC that provides modulation and demodulation functions and enables controller 536 to communicate via a wireless network. Network interface 524 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.

[0120] Vehicle 500 may further include data storage 528, which may include off-chip (e.g., off-chip SoC 504) storage devices. Data storage 528 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.

[0121] Vehicle 500 may further include GNSS sensor 558. GNSS sensor 558 (e.g., GPS, assisted GPS sensor, differential GPS (DGPS) sensor, etc.) is used for assisted mapping, sensing, occupancy grid generation, and / or path planning functions. Any number of GNSS sensors 558 can be used, including, for example, but not limited to, GPS using a USB connector with an Ethernet-to-serial (RS-232) bridge.

[0122] Vehicle 500 may further include a RADAR sensor 560. The RADAR sensor 560 can be used by vehicle 500 for remote vehicle detection even in dark and / or inclement weather conditions. The RADAR functional safety level may be ASIL B. The RADAR sensor 560 can use CAN and / or bus 502 (e.g., to transmit data generated by the RADAR sensor 560) 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 560 can be adapted for front, rear, and side RADAR use. In some examples, a pulse Doppler RADAR sensor is used.

[0123] The RADAR sensor 560 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, the long-range RADAR can be used for adaptive cruise control functions. The long-range RADAR system can provide a wide field of view (e.g., within 250m) achieved through two or more independent scans. The RADAR sensor 560 can help distinguish between stationary and moving objects and can be used by ADAS systems for emergency braking assist and forward collision warning. The long-range RADAR sensor 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 500'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 500's lane.

[0124] As an example, a mid-range RADAR system can include a range of up to 560m (front) or 80m (rear) and a field of view of up to 42 degrees (front) or 550 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.

[0125] Short-range RADAR systems can be used in ADAS systems for blind spot detection and / or lane change assistance.

[0126] Vehicle 500 may further include ultrasonic sensors 562. Ultrasonic sensors 562, which may be positioned at the front, rear, and / or sides of vehicle 500, can be used for parking assistance and / or creating and updating occupancy grids. A wide variety of ultrasonic sensors 562 can be used, and different ultrasonic sensors 562 can be used for different detection ranges (e.g., 2.5m, 4m). Ultrasonic sensors 562 can operate at functional safety level ASIL B.

[0127] Vehicle 500 may include a LiDAR sensor 564. The LiDAR sensor 564 may be used for object and pedestrian detection, emergency braking, collision avoidance, and / or other functions. The LiDAR sensor 564 may be of functional safety level ASIL B. In some examples, vehicle 500 may include multiple LiDAR sensors 564 (e.g., two, four, six, etc.) that can use Ethernet (e.g., to provide data to a Gigabit Ethernet switch).

[0128] In some examples, the LiDAR sensor 564 may be able to provide a list of objects and their distances within a 360-degree field of view. Commercially available LiDAR sensors 564 may have an advertising range of, for example, approximately 500m, with an accuracy of 2cm-3cm, and support for 500Mbps Ethernet connectivity. In some examples, one or more non-protruding LiDAR sensors 564 may be used. In such examples, the LiDAR sensor 564 may be implemented as a small device that can be embedded in the front, rear, sides, and / or corners of a vehicle 500. In such examples, the LiDAR sensor 564 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 564 may be configured for a horizontal field of view between 45 degrees and 135 degrees.

[0129] In some examples, LiDAR technologies such as 3D flash LiDAR can also be used. 3D flash LiDAR uses flashes 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-browsing LiDAR devices) without moving parts other than fans. 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 flash LiDAR, and because flash LiDAR is a solid-state device with no moving parts, the LiDAR sensor 564 is less susceptible to motion blur, vibration, and / or shock.

[0130] The vehicle may further include an IMU sensor 566. In some examples, the IMU sensor 566 may be located at the center of the rear axle of the vehicle 500. The IMU sensor 566 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 566 may include an accelerometer and a gyroscope, while in a nine-axis application, the IMU sensor 566 may include an accelerometer, a gyroscope, and a magnetometer.

[0131] In some embodiments, the IMU sensor 566 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 566 can enable the vehicle 500 to estimate heading by directly observing and correlating velocity changes from GPS to the IMU sensor 566 without input from a magnetic sensor. In some examples, the IMU sensor 566 and the GNSS sensor 558 can be combined into a single integrated unit.

[0132] The vehicle may include a microphone 596 placed in and / or around the vehicle 500. Among other things, the microphone 596 may be used for emergency vehicle detection and identification.

[0133] The vehicle may further include any number of camera types, including stereo camera 568, wide-angle camera 570, infrared camera 572, surround camera 574, long-range and / or mid-range camera 598, and / or other camera types. These cameras can be used to capture image data around the entire perimeter of the vehicle 500. The camera types used depend on the embodiment and the requirements of the vehicle 500, and any combination of camera types can be used to provide the necessary coverage around the vehicle 500. 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 5A and Figure 5B It was described in more detail.

[0134] Vehicle 500 may further include vibration sensor 542. Vibration sensor 542 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 542 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).

[0135] Vehicle 500 may include ADAS system 538. In some examples, ADAS system 538 may include SoC. ADAS system 538 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.

[0136] The ACC system can use a RADAR sensor 560, a LiDAR sensor 564, and / or a camera. The ACC system can include longitudinal ACC and / or lateral ACC. Longitudinal ACC monitors and controls the distance to vehicles immediately in front of vehicle 500 and automatically adjusts the vehicle speed to maintain a safe distance. Lateral ACC performs distance holding and, if necessary, advises vehicle 500 to change lanes. Lateral ACC is associated with other ADAS applications such as LCA and CWS.

[0137] CACC uses information from other vehicles, which can be received indirectly from other vehicles via a wireless link or network connection (e.g., via the Internet) through network interface 524 and / or wireless antenna 526. 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 500 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 500, CACC can be more reliable, and it has the potential to improve traffic flow and reduce road congestion.

[0138] The Forward Collision 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 560 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.

[0139] An AEB (Autonomous Emergency Braking) 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 560 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.

[0140] 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.

[0141] The LKA system is a variation of the LDW system. If vehicle 500 begins to leave the lane, the LKA system provides corrective steering input or braking to vehicle 500.

[0142] 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 utilize a rear-facing camera and / or RADAR sensor 560 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.

[0143] RCTW systems can provide visual, auditory, and / or tactile notifications when an object is detected outside the range of a rear-view camera while the vehicle is reversing. Some RCTW systems include AEB (Autonomous Emergency Braking) to ensure the application of the vehicle's brakes to avoid a collision. RCTW systems may use one or more rear-view RADAR sensors 560 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.

[0144] 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 an autonomous vehicle 500, in the event of conflicting results, the vehicle 500 itself must decide whether to heed the results from the main computer or auxiliary computer (e.g., the first controller 536 or the second controller 536). For example, in some embodiments, the ADAS system 538 may be a backup and / or auxiliary computer used to provide 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 538 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.

[0145] 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.

[0146] 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 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 components of SoC 504 and / or be included as components of SoC 504.

[0147] In other examples, ADAS system 538 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.

[0148] In some examples, the output of ADAS system 538 can be fed to the perception block and / or the dynamic driving task block of the main computer. For example, if ADAS system 538 issues a forward collision warning because an object is immediately in front, the perception block can use this information when recognizing 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.

[0149] Vehicle 500 may further include an infotainment SoC 530 (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 530 may include a combination of hardware and software that can be used to provide vehicle 500 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, Wi-Fi, 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 530 may include a radio, disc player, navigation system, video player, USB and Bluetooth connectivity, in-vehicle computer, in-vehicle entertainment, Wi-Fi, steering wheel audio controls, hands-free voice controls, head-up display (HUD), HMI display 534, 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 530 may further be used to provide information (e.g., visual and / or auditory) to the vehicle's users, such as information from the ADAS system 538, 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.

[0150] The infotainment SoC 530 may include GPU functionality. The infotainment SoC 530 can communicate with other devices, systems, and / or components of the vehicle 500 via bus 502 (e.g., CAN bus, Ethernet, etc.). In some examples, the infotainment SoC 530 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 536 (e.g., the primary and / or backup computer of the vehicle 500). In such an example, the infotainment SoC 530 may place the vehicle 500 into a driver-safe parking mode as described herein.

[0151] Vehicle 500 may further include an instrument cluster 532 (e.g., a digital instrument panel, electronic instrument cluster, digital instrument panel, etc.). The instrument cluster 532 may include a controller and / or a supercomputer (e.g., a discrete controller or supercomputer). The instrument cluster 532 may include a set of instruments such as a 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 the infotainment SoC 530 and the instrument cluster 532. In other words, the instrument cluster 532 may be included as part of the infotainment SoC 530, or vice versa.

[0152] Figure 5D For cloud-based servers and according to some embodiments of this disclosure Figure 5A This is a system diagram illustrating communication between example autonomous vehicles 500. System 576 may include server 578, network 590, and vehicles including vehicle 500. Server 578 may include multiple GPUs 584(A)-584(H) (collectively referred to herein as GPU 584), PCIe switches 582(A)-582(H) (collectively referred to herein as PCIe switch 582), and / or CPUs 580(A)-580(B) (collectively referred to herein as CPU 580). GPUs 584, CPUs 580, and PCIe switches may be interconnected with high-speed interconnects and / or PCIe connections 586, such as, but not limited to, NVLink interfaces 588 developed by NVIDIA. In some examples, GPUs 584 are connected via NVLink and / or NVSwitch SoCs, and GPUs 584 and PCIe switches 582 are connected via PCIe interconnects. Although eight GPUs 584, two CPUs 580, and two PCIe switches are shown in the diagram, this is not intended to be limiting. Depending on the embodiment, each of the servers 578 may include any number of GPUs 584, CPUs 580, and / or PCIe switches. For example, each of the servers 578 may include eight, sixteen, thirty-two, and / or more GPUs 584.

[0153] Server 578 can receive image data from vehicles via network 590, representing images of unexpected or changed road conditions such as recently commenced roadworks. Server 578 can also transmit neural network 592, updated neural network 592, and / or map information 594, including information about traffic and road conditions, to vehicles via network 590. Updates to map information 594 may include updates to HD map 522, such as information about construction sites, potholes, bends, floods, or other obstacles. In some examples, neural network 592, updated neural network 592, and / or map information 594 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 578 and / or other servers).

[0154] Server 578 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 analysis 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 590), and / or the machine learning model can be used by server 578 to remotely monitor the vehicle.

[0155] In some examples, server 578 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 578 may include a deep learning supercomputer powered by GPU 584 and / or a dedicated AI computer, such as the DGX and DGX Station machines developed by NVIDIA. However, in some examples, server 578 may include a deep learning infrastructure in a data center that uses only CPU power.

[0156] The deep learning infrastructure of server 578 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 500. For example, the deep learning infrastructure may receive periodic updates from vehicle 500, such as image sequences and / or objects located in those image sequences by vehicle 500 (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 500. If the results do not match and the infrastructure concludes that the AI ​​in vehicle 500 has malfunctioned, then server 578 may transmit a signal to vehicle 500 instructing its fail-safe computer to take control, notify passengers, and complete a safe stopping operation.

[0157] For inference, server 578 may include GPU 584 and one or more programmable inference accelerators (such as NVIDIA's TensorRT 3). 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.

[0158] Example computing device

[0159] Figure 6 The block diagram is provided for an example computing device 600 suitable for implementing some embodiments of the present disclosure. The computing device 600 may include an interconnect system 602 directly or indirectly coupled to the following devices: memory 604, one or more central processing units (CPUs) 606, one or more graphics processing units (GPUs) 608, a communication interface 610, input / output (I / O) ports 612, input / output components 614, a power supply 616, one or more presentation components 618 (e.g., a display), and one or more logic units 620. In at least one embodiment, the computing device 600 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 608 may include one or more vGPUs, one or more CPUs 606 may include one or more vCPUs, and / or one or more logic units 620 may include one or more virtual logic units. Therefore, computing device 600 may include discrete components (e.g., a complete GPU dedicated to computing device 600), virtual components (e.g., a portion of the GPU dedicated to computing device 600), or a combination thereof.

[0160] although Figure 6The various boxes are shown connected via an interconnect system 602 with wiring, but this is not intended to be limiting and is merely for clarity. For example, in some embodiments, a presentation component 618, such as a display device, may be considered an I / O component 614 (e.g., if the display is a touchscreen). As another example, CPU 606 and / or GPU 608 may include memory (e.g., memory 604 may represent a storage device other than the memory of GPU 608, CPU 606, and / or other components). In other words, Figure 6 The computing devices mentioned are merely illustrative. No distinction is made between categories such as "workstation," "server," "laptop," "desktop," "tablet," "client device," "mobile device," "handheld device," "game console," "electronic control unit (ECU)," "virtual reality system," and / or other device or system types, as all of these are considered within the same category. Figure 6 Within the scope of computing devices.

[0161] Interconnect system 602 may represent one or more links or buses, such as address buses, data buses, control buses, or combinations thereof. Interconnect system 602 may include one or more link or bus 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, Peripheral Component Interconnect Fast (PCIe) bus, and / or another type of bus or link. In some embodiments, there is a direct connection between components. As an example, CPU 606 may be directly connected to memory 604. Furthermore, CPU 606 may be directly connected to GPU 608. In cases where there is a direct or point-to-point connection between components, interconnect system 602 may include a PCIe link to perform the connection. In these examples, a PCI bus is not required in computing device 600.

[0162] Memory 604 may include any medium of a wide variety of computer-readable media. Computer-readable media can be any available medium that can be accessed by computing device 600. Computer-readable media may include volatile and non-volatile media, as well as removable and non-removable media. For example and without limitation, computer-readable media may include computer storage media and communication media.

[0163] Computer storage media may include volatile and non-volatile media and / or removable and non-removable media, implemented in any way or by any method or technique for storing information such as computer-readable instructions, data structures, program modules, and / or other data types. For example, memory 604 may store computer-readable instructions (e.g., representing programs and / or program elements, such as an operating system). Computer storage media may include, but is not limited to, RAM, ROM, EEPROM, flash memory or other storage technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage devices, magnetic tape cassettes, magnetic tape, disk storage devices or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by computing device 600. As used herein, computer storage media does not include the signal itself.

[0164] Computer storage media may contain 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 include any information transport medium. The term "modulated data signal" can refer to a signal whose characteristics are set or altered in a manner that encodes information into that signal. For example and without limitation, computer storage media may include wired media such as wired networks or direct wired connections, and wireless media such as sound, RF, infrared, and other wireless media. Any combination of the above should also be included within the scope of computer-readable media.

[0165] CPU 606 may be configured to execute at least some of computer-readable instructions to control one or more components of computing device 600 to perform one or more of the methods and / or processes described herein. Each of CPU 606 may include one or more cores (e.g., one, two, four, eight, twenty-eight, seventy-two, etc.) capable of processing a large number of software threads simultaneously. CPU 606 may include any type of processor and may include different types of processors depending on the type of computing device 600 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 600, the processor may be an advanced RISC mechanism (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 600 may also include one or more CPUs 606.

[0166] In addition to or replacing CPU 606, GPU 608 may also be configured to execute at least some computer-readable instructions to control one or more components of computing device 600 to perform one or more of the methods and / or processes described herein. One or more GPUs 608 may be integrated GPUs (e.g., having one or more CPUs 606) and / or one or more GPUs 608 may be discrete GPUs. In embodiments, one or more GPUs 608 may be coprocessors of one or more CPUs 606. Computing device 600 may use GPU 608 to render graphics (e.g., 3D graphics) or perform general-purpose computing. For example, GPU 608 may be used for general-purpose computing on a GPU (GPGPU). GPU 608 may include hundreds or thousands of cores capable of processing hundreds or thousands of software threads simultaneously. GPU 608 may generate pixel data for outputting an image in response to rendering commands (e.g., rendering commands received via a host interface from CPU 606). GPU 608 may include graphics memory, such as display memory, for storing pixel data or any other suitable data (e.g., GPGPU data). Display memory may be included as part of memory 604. GPU 608 may include two or more GPUs operating in parallel (e.g., via links). The links may connect the GPUs directly (e.g., using NVLINK) or via a switch (e.g., using NVSwitch). When combined, each GPU 608 may generate different portions of pixel data or GPGPU data for different outputs (e.g., the first GPU for the first image, the second GPU for the second image). Each GPU may include its own memory or may share memory with other GPUs.

[0167] In addition to or replacing CPU 606 and / or GPU 608, logic unit 620 may be configured to execute at least some computer-readable instructions to control one or more components of computing device 600 to perform one or more methods and / or processes described herein. In embodiments, CPU 606, GPU 608, and / or logic unit 620 may execute any combination of methods, processes, and / or portions thereof discretely or jointly. One or more logic units 620 may be part of and / or integrated into one or more CPUs 606 and / or one or more GPUs 608, and / or one or more logic units 620 may be discrete components of CPU 606 and / or GPU 608 or otherwise external thereto. In embodiments, one or more logic units 620 may be processors of one or more CPUs 606 and / or one or more GPUs 608.

[0168] Examples of logic unit 620 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 traversal 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 peripheral component interconnect fast (PCIe) element, etc.

[0169] The communication interface 610 may include one or more receivers, transmitters, and / or transceivers that enable the computing device 600 to communicate with other computing devices via electronic communication networks, including wired and / or wireless communications. The communication interface 610 may include components and functions that enable communication via any of several different networks, such as wireless networks (e.g., Wi-Fi, Z-Wave, Bluetooth, Bluetooth LE, ZigBee, etc.), wired networks (e.g., communication via Ethernet or InfiniBand), low-power wide area networks (e.g., LoRaWAN, SigFox, etc.), and / or the Internet. In one or more embodiments, one or more logic units 620 and / or the communication interface 610 may include one or more data processing units (DPUs) for directly transmitting data received via a network and / or via interconnect system 602 to one or more GPUs 608 (e.g., their memory).

[0170] I / O port 612 enables computing device 600 to be logically coupled to other devices, including I / O component 614, presentation component 618, and / or other components, some of which may be built into (e.g., integrated into) computing device 600. Illustrative I / O component 614 includes microphones, mice, keyboards, joysticks, game pads, game controllers, satellite dish antennas, browsers, printers, wireless devices, and so on. I / O component 614 can provide a Natural User Interface (NUI) for processing user-generated air gestures, voice, or other physiological input. In some instances, the input may be transmitted to appropriate network elements for further processing. The NUI can implement any combination of voice recognition, stylus recognition, facial recognition, biometric recognition, on-screen and adjacent-screen gesture recognition, air gestures, head and eye tracking, and touch recognition associated with the display of computing device 600 (described in more detail below). Computing device 600 may include depth cameras such as stereo camera systems, infrared camera systems, RGB camera systems, touchscreen technology, and combinations thereof for gesture detection and recognition. In addition, computing device 600 may include an accelerometer or gyroscope that enables motion detection (e.g., as part of an inertial measurement unit (IMU)). In some examples, the output of the accelerometer or gyroscope may be used by computing device 600 to render immersive augmented reality or virtual reality.

[0171] Power supply 616 may include a hard-wired power supply, a battery power supply, or a combination thereof. Power supply 616 may supply power to computing device 600 so that components of computing device 600 can operate.

[0172] The presentation component 618 may include a display (such as a monitor, touch screen, television screen, head-up display (HUD), other display types, or combinations thereof), speakers, and / or other presentation components. The presentation component 618 may receive data from other components (such as GPU 608, CPU 606, DPU, etc.) and output that data (e.g., as an image, video, sound, etc.).

[0173] Example Data Center

[0174] Figure 7 An example data center 700 is shown, which can be used in at least one embodiment of this disclosure. The data center 700 may include a data center infrastructure layer 710, a framework layer 720, a software layer 730, and an application layer 740.

[0175] like Figure 7As shown, the data center infrastructure layer 710 may include a resource coordinator 712, grouped computing resources 714, and node computing resources (“nodes CR”) 716(1)-716(N), where “N” represents any complete positive integer. In at least one embodiment, nodes CR 716(1)-716(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 drives or disk drives), network input / output (NW I / O) devices, network switches, virtual machines (VMs), power modules, and cooling modules, etc. In some embodiments, one or more nodes CR 716(1)-716(N) may correspond to servers having one or more of the aforementioned computing resources. In addition, in some embodiments, nodes CR716(1)-716(N) may include one or more virtual components, such as vGPU, vCPU, etc., and / or one or more of nodes CR716(1)-716(N) may correspond to virtual machines (VMs).

[0176] In at least one embodiment, the grouped computing resources 714 may include individual groups (not shown) of nodes CR716 housed in one or more racks, or a plurality of racks (also not shown) housed in data centers in various geographic locations. Individual groups of nodes CR716 within the grouped computing resources 714 may include computing, networking, memory, or storage resources that can be configured or allocated to support groups of one or more workloads. In at least one embodiment, several nodes CR716, 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.

[0177] Resource coordinator 712 can be configured or otherwise controlled to control one or more nodes CR716(1)-716(N) and / or groups of computing resources 714. In at least one embodiment, resource coordinator 712 may include a Software Design Infrastructure (SDI) management entity for data center 700. Resource coordinator 712 may include hardware, software, or some combination thereof.

[0178] In at least one embodiment, such as Figure 7As shown, framework layer 720 may include a job scheduler 732, a configuration manager 734, a resource manager 736, and a distributed file system 738. Framework layer 720 may include a framework of software 732 supporting software layer 730 and / or one or more applications 742 of application layer 740. Software 732 or application 742 may respectively include web-based service software or applications, such as service software or applications provided by Amazon Web Services, Google Cloud, and Microsoft Azure. Framework layer 720 may be, but is not limited to, a free and open-source software web application framework, such as Apache Spark, which can utilize distributed file system 738 for large-scale data processing (e.g., "big data"). TM (Hereinafter referred to as "Spark"). In at least one embodiment, the job scheduler 732 may include a Spark driver for facilitating the scheduling of workloads supported by various layers of data center 700. In at least one embodiment, the configuration manager 734 may be able to configure different layers, such as software layer 730 and framework layer 720 including Spark and a distributed file system 738 for supporting large-scale data processing. The resource manager 736 is able to manage cluster or grouped computing resources mapped to or allocated for supporting distributed file system 738 and job scheduler 732. In at least one embodiment, cluster or grouped computing resources may include grouped computing resources 714 at data center infrastructure layer 710. The resource manager 736 may coordinate with resource coordinator 712 to manage these mapped or allocated computing resources.

[0179] In at least one embodiment, the software 732 included in the software layer 730 may include software used by at least a portion of the nodes CR716(1)-716(N), the grouped computing resources 714, and / or the distributed file system 738 of the framework layer 720. One or more types of software may include, but are not limited to, Internet web page search software, email virus browsing software, database software, and streaming video content software.

[0180] In at least one embodiment, the application layer 740 may include one or more applications 742 that can be used by at least a portion of the nodes CR716(1)-716(N), the grouped computing resources 714, and / or the distributed file system 738 of the framework layer 720. The 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 conjunction with one or more embodiments.

[0181] In at least one embodiment, any of the configuration manager 734, resource manager 736, and resource coordinator 712 can perform 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 alleviate the risk of data center operators of data center 700 making potentially poor configuration decisions and can prevent underutilization and / or skewed portions of the data center.

[0182] Data center 700 may include tools, services, software, or other resources for training one or more machine learning models or using one or more machine learning models to predict or infer information according to one or more embodiments described herein. For example, a machine learning model can be trained by calculating weight parameters based on a neural network architecture using the software and computing resources described above with respect to data center 700. In at least one embodiment, by using weight parameters calculated through one or more training techniques, information can be inferred or predicted using trained machine learning models corresponding to one or more neural networks, such as, but not limited to, those described herein, using the resources described above with respect to data center 700.

[0183] In at least one embodiment, the data center 700 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. Furthermore, one or more of the aforementioned software and / or hardware resources may be configured as a service to allow a user to train or perform information inference, such as image recognition, speech recognition, or other artificial intelligence services.

[0184] Example network environment

[0185] 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... Figure 6 The implementation is carried out on one or more instances of computing device 600—for example, each device may include similar components, features, and / or functions of computing device 600. Furthermore, in the case of implementing back-end devices (e.g., servers, NAS, etc.), the back-end devices may be included as part of data center 700, examples of which are described herein. Figure 7 To describe in more detail.

[0186] 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 networks within 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. In cases 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.

[0187] 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 server functionality described herein can be implemented on any number of client devices.

[0188] In at least one embodiment, the network environment may include one or more cloud-based network environments, distributed computing environments, combinations thereof, etc. A 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 for supporting software at the software layer and / or one or more applications at the application layer. The software or applications 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 type of free and open-source software network application framework, such as one that can use a distributed file system for large-scale data processing (e.g., "big data").

[0189] 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 various 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.). If the connection to a user (e.g., a client device) is relatively close to an edge server, the core server can assign at least a portion of the functionality 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).

[0190] Client devices may include those described in this article. Figure 6 The example computing device 600 described includes at least some components, features, and functions. By way of example and not limitation, the client device may be 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, aircraft, virtual machine, drone, robot, handheld communication device, hospital equipment, gaming equipment or system, entertainment system, in-vehicle computer system, embedded system controller, remote control, electrical appliance, consumer electronics device, workstation, edge device, any combination of these described devices, or any other suitable device.

[0191] 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.

[0192] 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" could 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 elements A, B, and C. Furthermore, "at least one of element A or element B" could 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" could 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.

[0193] This document describes in detail the subject matter of this disclosure to satisfy legal requirements. However, the description itself is not intended to limit the scope of this disclosure. Rather, the inventors have envisioned that the claimed subject matter may also 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 "block" may be used herein to imply different elements of the method employed, these terms should not be construed as implying any particular order among or between the various steps disclosed herein, unless the order of the steps is explicitly described.

Claims

1. A processor comprising one or more circuits, said one or more circuits being used to: Using sensor data obtained from one or more sensors of the self-machine, a first distance image corresponding to a first time point and a second distance image corresponding to a second time point following the first time point are generated; A first projected image is generated based at least on a first depth value from the first distance image projected onto a first coordinate space of the second distance image, and a second projected image is generated based at least on a second depth value from the second distance image projected onto a second coordinate space of the first distance image; The input data of the deep neural network (DNN) is calculated based on the first projected image and the second projected image. The input data indicates the depth difference between the first distance image and the second distance image, as well as the spatial correspondence difference of scene points between the first coordinate space and the second coordinate space. as well as The input data is processed at least based on the DNN to calculate output data indicating motion at one or more pixels in the second distance image.

2. The processor of claim 1, further comprising one or more circuits, said one or more circuits being used to: A comparison image is generated, at least in part, based on comparing a second depth value of the second distance image in the second coordinate space of the first distance image with a depth value of the first distance image; and The input data is calculated based on the comparison images.

3. The processor of claim 1, wherein the input data includes a first representation of one or more regions of the first projected image and a second representation of one or more regions of the second projected image.

4. The processor of claim 1, wherein the first projected image and the second projected image are generated using the self-motion of the self-machine tracked between the first time and the second time.

5. The processor of claim 1, wherein the input data includes a first channel and a second channel, the first channel corresponding to a first distance between the self-machine at a first position at the first time and a first 3D point, the second channel corresponding to a second distance between the self-machine at a second position at the second time and a second 3D point, wherein when viewed from the self-machine at the second position at the second time, the first 3D point and the second 3D point are projected onto the same pixel.

6. The processor of claim 1, wherein the output data computed using the DNN represents a motion mask, the motion mask having one or more first values ​​indicating that one or more first objects described using the second distance image are in motion at the second time, and the motion mask having one or more second values ​​indicating that one or more second objects described using the second distance image are static at the second time.

7. The processor of claim 6, wherein for each pixel of one or more pixels in the second distance image, the motion mask includes a confidence value between 0 and 1, the confidence value indicating whether a feature or object represented by the corresponding pixel in the one or more pixels corresponds to an object in motion at the second time.

8. The processor of claim 1, wherein the one or more circuits are further configured to use the output data indicating motion to perform one or more control operations corresponding to the ego machine, at least based on motion at the one or more pixels of the second distance map.

9. The processor of claim 1, wherein the DNN is trained to predict the probability that one or more pixels correspond to a static or dynamic object.

10. The processor of claim 1, wherein the processor 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 used to perform simulation operations; A system used to perform deep learning operations; Systems implemented using edge devices; Systems implemented using robots; 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.

11. A method comprising: The self-generated machine's sensors generate a first distance image at a first moment and a second distance image at a second moment following the first moment; A first projected image is generated based at least on projecting one or more first depth values ​​from the first distance image onto a first coordinate space of the second distance image; A second projected image is generated based at least on projecting one or more second depth values ​​from the second distance image onto the second coordinate space of the first distance image; The input data of the deep neural network (DNN) is calculated based on the first projected image and the second projected image. The input data indicates the depth difference between the first distance image and the second distance image, as well as the spatial correspondence difference of scene points between the first coordinate space and the second coordinate space. as well as The input data is processed at least based on the DNN to calculate output data indicating motion at one or more pixels in the second distance image.

12. The method of claim 11, further comprising: A comparison image is generated, at least in part, based on comparing the second depth value of the second distance image in the second coordinate space of the first distance image with the depth value of the first distance image; The input data is calculated based on the comparison images.

13. The method of claim 11, wherein the input data includes a first representation of one or more regions of the first projected image and a second representation of one or more regions of the second projected image.

14. The method of claim 11, wherein the generation of the first projected image and the second projected image is based at least on the self-motion of the self-machine tracked between the first time and the second time.

15. A system comprising: One or more processors, said one or more processors being configured to perform operations, said operations including: The self-generated machine's sensors generate a first distance image at a first moment and a second distance image at a second moment following the first moment; A first projected image is generated based at least on projecting one or more first depth values ​​from the first distance image onto a first coordinate space of the second distance image; A second projected image is generated based at least on projecting one or more second depth values ​​from the second distance image onto the second coordinate space of the first distance image; The input data of the deep neural network (DNN) is calculated based on the first projected image and the second projected image, wherein the input data indicates the depth difference between the first distance image and the second distance image, and the spatial correspondence difference of scene points between the first coordinate space and the second coordinate space; and Using the DNN with the input data, output data indicating motion at one or more pixels in the second distance image is calculated.

16. The system of claim 15, wherein the operation further comprises: A comparison image is generated, at least in part, based on comparing the second depth value of the second distance image in the second coordinate space of the first distance image with the depth value of the first distance image. The input data is calculated based on the comparison images.

17. The system of claim 15, wherein for each of the n distance images preceding the second distance image, the second distance image is used to generate a corresponding first projection image, a second projection image, and a comparison image.

18. The system of claim 15, wherein the generation of the first projected image and the second projected image is based at least on the self-motion of the self-machine tracked between the first time and the second time.

19. The system of claim 15, wherein the operation further comprises: At least one or more object detections are computed based on the output data calculated using the DNN.

20. The system of claim 15, 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 used to perform simulation operations; A system used to perform deep learning operations; Systems implemented using edge devices; Systems implemented using robots; 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.

Images