Restraint Device Localization

Abstract

To provide systems and methods related to restraint device (e.g., seatbelt) localization.SOLUTION: In one embodiment, the disclosure relates to systems and methods for seatbelt detection and modeling. A vehicle may be occupied by one or more occupants wearing one or more seatbelts. A camera or other sensor is placed within the vehicle to capture images of the one or more occupants. A system analyzes the images to detect and model the seatbelts depicted in the images. Specifically, the system may scan the images and areas of the images that may correspond to the seatbelts. The system may assemble candidate areas of the images that may correspond to the seatbelts, and refine the candidate areas based on various constraints. The system may build models based on the refined candidate areas that indicate the seatbelts. The system may visualize the models indicating the seatbelts using the images.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] This application incorporates by reference for all purposes the entire disclosures of co-pending U.S. patent application Ser. No. 17 / 005,914, filed Aug. 28, 2020, entitled "NEURAL NETWORK BASED DETERMINATION OF GAZE DIRECTION USING SPATIAL MODELS," co-pending U.S. patent application Ser. No. 17 / 004,252, filed Aug. 27, 2020, entitled "NEURAL NETWORK BASED FACIAL ANALYSIS USING FACIAL LANDMARKS AND ASSOCIATED CONFIDENCE VALUES," and co-pending U.S. patent application Ser. No. 16 / 905,418, filed June 18, 2020, entitled "MACHINE LEARNING-BASED SEATBELT DETECTION AND USAGE RECOGNITION USING FIDUCIAL MARKING." [Background technology]

[0002] Automobiles and other vehicles and machines that transport or contain passengers or operators often have various safety features, such as seat belts or other restraining devices. In many cases, safety devices can be less effective or even ineffective if not used correctly. For example, an incorrectly worn seat belt can be significantly less effective than a correctly worn seat belt. Various attempts have been made to improve the use of such safety devices. Some devices involve retrofitting machines, which can be difficult and expensive, especially considering variations between machines of the same type. In general, detecting the correct use of restraining devices such as seat belts in vehicles presents many challenges and has been met with varying success in terms of accuracy.

[0003] At least one embodiment relates to a processing resource used to identify and model one or more constraining devices from one or more images, for example, at least one embodiment relates to a processor or computer system used to identify and model one or more constraining devices from one or more images of the one or more constraining devices according to various novel techniques described herein.

[0004] The present systems and methods for localization of restraining devices are described in detail below with reference to the accompanying drawings. [Brief description of the drawings]

[0005] [Figure 1] FIG. 1 illustrates an example of a system for seat belt localization, according to some embodiments of the present disclosure. [Diagram 2] FIG. 1 illustrates an example of a seat belt patch of one or more pixels with one or more neighboring pixels in one or more target directions, according to some embodiments of the present disclosure. [Diagram 3] FIG. 1 illustrates an example of a non-seatbelt patch of one or more pixels with one or more neighboring pixels in one or more target directions, according to some embodiments of the present disclosure. [Figure 4] FIG. 1 illustrates an example of a seat belt patch geometry, according to some embodiments of the present disclosure. [Diagram 5] FIG. 2 illustrates an example of structured edge localization, according to some embodiments of the present disclosure. [Figure 6] FIG. 2 illustrates an example of a candidate map, according to some embodiments of the present disclosure. [Figure 7] FIG. 13 illustrates an example of multiple seat belt localization, according to some embodiments of the present disclosure. [Figure 8] FIG. 1 illustrates an example of seat belt localization, according to some embodiments of the present disclosure. [Figure 9]FIG. 1 illustrates an example of seat belt localization, according to some embodiments of the present disclosure. [Figure 10] FIG. 1 illustrates an example of seat belt localization, according to some embodiments of the present disclosure. [Figure 11] FIG. 1 illustrates an example of seat belt localization, according to some embodiments of the present disclosure. [Figure 12] 1 is a flow diagram illustrating a method for generating a model of an input image according to some embodiments of the present disclosure. [Figure 13A] 1 is an illustration of an example autonomous vehicle, according to some embodiments of the present disclosure. [Figure 13B] 13B is an illustration of camera positions and fields of view for the example autonomous vehicle of FIG. 13A, according to some embodiments of the present disclosure. [Figure 13C] FIG. 13B is a block diagram of an example system architecture of the example autonomous vehicle of FIG. 13A in accordance with some embodiments of the present disclosure. [Figure 13D] FIG. 13B is a system diagram of communication between a cloud-based server and the example autonomous vehicle of FIG. 13A, according to some embodiments of the present disclosure. [Figure 14] FIG. 1 is a block diagram of an example computing device suitable for use in implementing some embodiments of the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0006] The techniques described herein provide a method to passively detect seat belts and other restraint devices to achieve various functions, such as detecting whether a vehicle occupant is wearing a seat belt and whether it is properly worn. A camera installed in the vehicle captures images, which are analyzed to determine whether the occupant is properly wearing the seat belt. Typically, the vehicle's system uses a sensory system to detect whether the seat belt is in a locked position. However, there are cases where the sensory system can be deceived or permanently non-functional. In addition, there are cases where the occupant is wearing the seat belt but is wearing it incorrectly (e.g., the seat belt is worn behind the back). To combat this and improve road safety, the techniques described herein use videos / images captured by a camera mounted on the vehicle to determine whether the seat belt is properly worn. The techniques are performed without the need to physically modify the seat belt itself or any existing locking sensory system.

[0007] More specifically, the system first analyzes the captured image to classify whether a pixel is part of a seat belt or part of the background. This classification results in the generation of multiple pixel candidates that are most likely to be part of a seat belt. To more accurately identify pixels that are part of a seat belt, the system uses various information to correct misidentification of images identified as part of a seat belt and vice versa. Such information may include: parameters / constraints such as how the seat belt is constructed in the vehicle, the direction the seat belt should extend when locked, the physical attributes of the seat belt, and the configuration of the camera that captured the image to construct a more accurate set of pixels. After filtering the pixels to generate a more accurate set of pixels, the system then parameterizes the seat belt and models the shape of the seat belt using a high order polynomial curve, which further removes any pixels that are outliers with respect to the modeled shape and recovers pixels that were previously incorrectly filtered out or occluded. The final seat belt curve may be used to improve occupant safety, for example, to determine whether the seat belt is properly fastened (e.g., the seat belt is in the locked position and fastened diagonally across the occupant's chest).

[0008] The techniques described herein further develop a method for improving safety features in all types of vehicles without the need to modify existing vehicle systems. The techniques can identify which occupants in a vehicle are properly seat belted based on video / images captured by on-board systems and / or cameras. The system can not only identify whether the driver is properly seat belted, but can also determine whether all other occupants in the vehicle are properly seat belted. Although the system described herein applies to seat belts in general passenger vehicles, the system can also be applied to other fields requiring seat belts, safety devices, or other harnesses (e.g., construction equipment, amusement park rides, 4D movie theater seats). The system can be a component of an in-vehicle occupant monitoring system (OMS).

[0009] The system can be robust and applicable to various types of camera sensor configurations, e.g., color or infrared cameras, normal view or fish-eye cameras. The system can operate and process images captured in various light conditions, e.g., low light conditions (e.g., nighttime conditions), variable light conditions (e.g., daytime conditions), and / or variations thereof. The system can determine new local parallel line patterns and detect patterns at a micro-scale in images via one or more techniques described herein. The system can efficiently determine all seat belts simultaneously for all frames of a video using parallel computing techniques using one or more general-purpose graphics processing units (GPGPUs). The system can perform real-time monitoring of seat belts and seat belt usage to improve the safety of one or more occupants occupying a vehicle. The system can use various algorithms that can provide a binary classification function for any patch of an image and can filter out noise or false positive seat belt portion candidates. The system can use a seat belt shape modeling function based on a high-order curve that can also localize seat belts with occlusions or other obstructions. The system can localize a seat belt corresponding to a subject and can be applicable to subjects of any suitable physical appearance, including physical appearances of different sizes, shapes, and attributes, such as clothing type and hair type, and / or various other physical appearance attributes.

[0010] Referring to FIG. 1, FIG. 1 is an illustration of a system for seat belt localization according to some embodiments of the present disclosure. It should be understood that this and other arrangements described herein are described by way of example only. Other arrangements and elements (e.g., machines, interfaces, functions, sequences, groupings of functions, 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 may be implemented as separate or distributed components or in conjunction with other components, and in any suitable combination and location. Various functions described herein as being performed by entities may be implemented by hardware, firmware, and / or software. For example, various functions may be implemented by a processor executing instructions stored in a memory.

[0011] FIG. 1 illustrates an example system 100 for seat belt localization according to at least one embodiment. In various embodiments, seat belt localization refers to one or more processes that identify the location, size, shape, orientation, and / or other characteristics of one or more seat belts shown in one or more images (e.g., images of a person in a driver's or passenger's seat of a vehicle). It should be noted that while example 100 illustrates seat belt localization, any suitable restraining device portion of any suitable system may be localized. Restraining devices may include devices such as seat belts (e.g., two-point seat belts, three-point seat belts, four-point seat belts, etc.), safety harnesses, or other vehicle safety devices.

[0012] The system for seat belt localization, which may be referred to as a seat belt localizer, may include a local predictor 104, a global assembler 108, and a shape modeler 116. Various processes of the system for seat belt localization may be performed by one or more graphics processing units (GPUs), such as parallel processing units (PPUs). One or more processes of the system for seat belt localization may be performed by any suitable processing system or unit (e.g., GPUs, PPUs, central processing units (CPUs)) and in any suitable manner, including sequentially, in parallel, and / or variations thereof. The local predictor 104, the global assembler 108, and the shape modeler 116 may be software modules of one or more computer systems onboard the vehicle. In some instances, the local predictor 104, the global assembler 108, and the shape modeler 116 are software programs executing on a computer server accessible over one or more networks, where the input image 102 is provided to the computer server from one or more computing systems in the vehicle over one or more networks, and seat belt localization results of one or more seat belts in the input image 102 are provided back to the vehicle over one or more networks. An input to the system for seat belt localization may include the input image 102. In one embodiment, the input image 102 is an image of an entity in a vehicle having a restraining device, e.g., a seat belt. With reference to FIG. 1, the input image 102 may be an image of a person wearing a seat belt seated in the driver's seat of the vehicle. The system for seat belt localization may determine the position and orientation of the seat belt in the input image 102, and may further determine whether the seat belt in the input image 102 is correctly applied.

[0013] The input image 102 may be an image captured from one or more image capture devices, such as a camera or other device. The input image 102 may be a frame of video captured from one or more cameras. The input image 102 may be captured from one or more cameras located inside the vehicle. In some implementations, the input image 102 may be captured from one or more cameras outside the vehicle, such as via a surveillance system, a cell phone, a drone, a handheld imaging device, or other imaging system separate from the vehicle, capturing images of the vehicle and vehicle occupants. In some instances, the input image is data captured from an image capture device that is further processed by adjusting one or more color characteristics, upsampling or downsampling, cropping, and / or other processing of the data. The vehicle may be a vehicle, such as an autonomous vehicle, a semi-autonomous vehicle, a manual vehicle, and / or variations thereof. In some instances, the vehicle is a recreational vehicle (e.g., a roller coaster), a construction vehicle, or other vehicle that requires one or more restraining devices. The input image 102 may be captured from one or more cameras, such as those described with respect to FIGS. 13A-13D. The input image 102 may be captured from a camera in the vehicle facing the vehicle occupant. The input image 102 may be a color image, a grayscale image, a black and white image, etc. The input image 102 may be captured from one or more cameras with different FOVs. The input image 102 may be an image with minimal color contrast, which may result in low contrast between the seat belt and the background of the input image 102. The input image 102 may show a seat belt that may be occluded by one or more entities, such as a hand, arm, clothing, or other object. The input image 102 may show a seat belt that may be moved to various locations or stretched. The input image 102 may show a seat belt that may be positioned as a straight line or other irregular shape, such as a curve. The input image 102 may be distorted as a result of object or vehicle movement. The input image 102 may show a subject, for example a vehicle driver or passenger, and a restraining device, for example a seat belt, applied to the subject.In some instances, the input image 102 shows a subject and a restraining device in an environment, such as a constructed environment (eg, the input image 102 shows a subject in a constructed environment having a restraining device, such as a safety harness).

[0014] The input image 102 may be received or otherwise obtained by the local predictor 104. In some instances, the input image 102 is processed by one or more pre-processing operations to prepare the input image 102 for processing, which may include operations such as feature / contrast enhancement, noise removal, morphological operations, resizing / scaling, and / or modifications thereof. The local predictor 104 may be a collection of one or more hardware and / or software computational resources having executable instructions that, when executed, scan the input image and predict areas of the input image that are part of a seat belt. The input image 102 may include a collection of pixels, where a seat belt depicted in the input image 102 may occupy multiple pixels. For any given pixel, the local predictor 104 may use information from the pixel and its neighboring pixels to predict whether the pixel belongs to a seat belt or is a background pixel. Pixels determined to be part of a seat belt may be identified as part of the foreground of the input image 102, and pixels determined not to be part of a seat belt may be identified as part of the background of the input image 102. The local predictor 104 may perform an initial classification of areas of the input image 102 to determine areas of the input image 102 that depict or otherwise represent a restraining device (e.g., a seat belt). An area may include a pixel, a collection of pixels, etc. An area may include a grouping of pixels that may be in close proximity to one another. An area may include a contiguous group or region of pixels of the image.

[0015] The local predictor 104 may generate a true result, which may be indicated by a value of 1, or a false result, which may be indicated by a value of 0, representing a seat belt or not a seat belt, respectively. In various embodiments, the input image 102 includes noise, and the classification result may or may not be correct initially. In one embodiment, there are four possible cases, although there may be additional cases as well: (1) the pixel is part of a seat belt and the local predictor 104 returns a 1, true positive; (2) the pixel is not part of a seat belt and the local predictor 104 returns a 1, false positive; (3) the pixel is part of a seat belt and the local predictor 104 returns a 0, false negative; and (4) the pixel is not part of a seat belt and the local predictor 104 returns a 0, true negative. The local predictor 104 may prioritize high recall over high precision (e.g., the local predictor 104 may reduce false negatives despite the possibility of increased false positives).

[0016] The local predictor 104 may use nearby neighboring pixels for additional information. The pixel of the input image 102 located at the yth row and xth column of the input image 102 may be denoted as p(x,y), or any variation thereof. The dimensions of the input image 102 may be denoted as W width, H height, and K channels (e.g., K=3 (or 4) for color images, and K=1 for gray images). p(x,y) and its neighboring pixels may constitute a set S, which may be represented by the following equation, although any variation thereof may be used: S={p ij |f dist (p ij ,p)≦ε,0 <i≦W,0<j≦H}(1)、 So, f dist may be a pixel distance function (eg, L1 distance), and ε may be a threshold that determines the neighborhood boundary.

[0017] In some instances, the candidate neighborhood shapes, also referred to as patches, include circles, rectangles, squares, and / or variations thereof. In one embodiment, the local predictor 104 uses a square with the candidate pixel located at its center as the neighborhood shape. The length of the square may be denoted by L=2k+1, where k=1, 2, 3..., where the set S may be simplified via the following equation, although any variations thereof may be used: S={p ij |0≦|iy|≦k,0≦|jx|≦k} (2).

[0018] The local predictor 104 may perform isotropy checks along various directions. The local predictor 104 generates a binary prediction result vector, which may be represented by the following equation:

number

number

number

number

[0019] f criteria (x,y,θ i ) the criterion function is a function of the input image 102 and a specified direction θ iFor any pixel p(x,y) in f, it can be determined whether the patch image generated using p and its neighbors S in this direction is a seat belt component or not. predictor The local predictor, which may be denoted by (x,y), is

number

number

number

[0020] With reference to FIG. 2, the example 200 may include (a) a seat belt patch sample shown at the top left, (b) an expanded seat belt patch shown at the top right, (c) a seat belt patch projection curve shown at the bottom left, and (d) a smoothed seat belt curve shown at the bottom right. With reference to FIG. 2, the (c) seat belt patch projection curve may correspond to the (b) expanded seat belt patch, the x-axis of (c) may correspond to the x or horizontal value of (b), and the y-axis of (c) may correspond to the pixel intensity values ​​of the vertical pixel column of (b). The pixel intensity of a particular pixel column may be the sum of all pixel intensities of the pixels of the pixel column, also referred to as the pixel intensity level. The pixel intensity or pixel intensity level of a particular pixel may correspond to the brightness of the pixel (e.g., a pixel with a high intensity may appear white and a pixel with a low intensity may appear black). With reference to FIG. 2, the (d) smoothed seat belt curve may be a smoothed version of the (c) seat belt patch projection curve. For example, referring to Figure 2, the two peaks in (c) the seatbelt patch projection curve and (d) the smoothed seatbelt curve correspond to the white vertical lines in (b) the extended seatbelt patch, which may correspond to the seatbelt boundaries.

[0021] With reference to FIG. 3, the example 300 may include (e) a non-seatbelt patch sample shown in the upper left, (f) an extended non-seatbelt patch shown in the upper right, (g) a non-seatbelt patch projection curve shown in the lower left, and (h) a smoothed non-seatbelt curve shown in the lower right. With reference to FIG. 3, (g) the non-seatbelt patch projection curve may correspond to (f) an extended non-seatbelt patch, where the x-axis of (g) may correspond to the x or horizontal value of (f) and the y-axis of (g) may correspond to the pixel intensity values ​​of a vertical pixel column of (f). The pixel intensity of a particular pixel column may be the sum of all pixel intensities of the pixels in the pixel column. With reference to FIG. 3, (h) the smoothed non-seatbelt curve may be a smoothed version of (f) the extended non-seatbelt patch.

[0022] 2 and 3 show examples of potential seat belt and non-seat belt patches / curves, it should be noted that the seat belt and / or non-seat belt patches / curves may be any variations thereof. A seat belt patch sample may correspond to any patch or region of an image (e.g., input image 102) that may show one or more seat belts, and the corresponding seat belt patch curves may have any suitable shape based at least in part on the seat belt patch sample.

[0023] Various information can be determined from Figures 2 and 3. In one case, a structured edge is observed because the edge of the seat belt may generate two approximately parallel lines in the patch direction based on the scale of the patch (e.g., if the patch is small enough), where the structured edge can be parameterized in terms of the distance range between the edges, and variations thereof, based on the relative camera-seat belt geometric relationship.

[0024] Since seat belts can be of various colors, e.g., black, gray, tan, etc., intensity and / or saturation regions can also be observed, where pixel values ​​can change with various environmental changes (e.g., lighting), but seat belt pixel values ​​can have several intensity ranges (e.g., seat belt pixels are rarely pure white).

[0025] In one embodiment, surface smoothness is also observed since a seat belt may have a similar texture along the belt, so pixels within different seat belt regions may have limited diversity in terms of smoothness.

[0026] In one embodiment, f structure , f intensity , and f smoothness can be expressed as a binary function as follows: f criteria (x,y,θ i )=f structure (x,y,θ i )∩f intensity (x,y,θi )∩f smoothness (x,y,θ i ) (6). Structural standard f structure may be defined as how it represents the parallel edges of a seat belt boundary line within a pixel's neighborhood. The local predictor 104 may identify and localize candidate seat belt edges and use them to distinguish seat belt pixels from non-seat belt pixels (e.g., noise).

[0027] A seat belt patch may have one or more orientations, and for each patch, the local predictor 104 may determine whether a seat belt boundary exists along the patch orientation. The patch orientations may be equally spaced into D categories in the range [0,π) (e.g., as shown in equation (3)), where regardless of how the seat belt is pointed, its orientation may be classified into one of the categories. For a given patch orientation θ i For i The local predictor 104 can determine seat belt patches in any suitable direction and seat belt boundaries along the direction of the seat belt.

[0028] FIG. 4 illustrates an example seat belt patch geometry 400 according to at least one embodiment. The seat belt patch geometry may refer to one or more geometry or geometric features of a particular seat belt patch, such as patch direction, angle, orientation, size, perimeter, etc. It should be noted that FIG. 4 illustrates an example of potential seat belt patch geometry, and the seat belt patch geometry may be any variation thereof. The seat belt patch may correspond to any patch or region of an image (e.g., input image 102) that may show one or more seat belts, and the corresponding seat belt patch geometry may have any suitable geometry based at least in part on the shape or geometry of the seat belt patch. With reference to FIG. 4, for each pixel (i,j) in the patch generated around pixel p(x,y), a value from pixel (x',y') may be obtained from the image (e.g., input image 102) using the following formula, although any variation thereof may be used:

number

[0029] The local predictor 104 may obtain the seat belt patch and / or the non-seat belt patch using one or more equations, for example, equation (7) as described above. The local predictor 104 may determine whether the patch has a pair of structured edges. The edge of the patch may be one or more successions of aligned pixel rows with an abrupt pixel intensity change at the same position within each row. The change may be a pixel intensity increment or decrement. The index of the abrupt change may be where the line may be located. In some instances, in one case of two parallel lines, the patch has a structured edge.

[0030] The local predictor 104 can use two-dimensional (2D) gradient operations to capture the magnitude of sudden intensity changes and accumulate them along the patch direction to distinguish seat belt ends from background noise. patch (x,y,θ,j) may be defined as the resulting curve after projection and may be represented by the following equation, although any variation thereof may be used:

number

[0031] The curve may be subject to various perturbations, for example, changes in illumination, as well as other noises. The local predictor 104 may apply an algorithm, such as the Savtzsky-Golay algorithm, to smooth the curve for further processing. patch (x,y,θ,j) may be defined as the filtered curve and may be represented by the following equation, although any variation thereof may be used:

number

[0032] Seatbelt patches with structured edges may have a filtered curve pattern similar to that shown in Fig. 5, while non-seatbelt patches may have random curves. The local predictor 104 extracts various characteristics from the curves to estimate the seatbelt pattern, e.g., the distance between the edges d edges and peak value difference d peaks In one embodiment, f structure (x,y,θi ) is defined by the following formula, although any variation thereof may be used:

number

[0033] FIG. 5 illustrates an example structured edge localization 500 according to at least one embodiment. The example 500 illustrates determining the edges of a seat belt from a seat belt patch curve, such as those shown in FIGS. 2 and 3. It should be noted that FIG. 5 illustrates one example of a potential localization of a structured edge from a seat belt patch curve, and that the localization of a structured edge from a seat belt patch curve may be any variation thereof. The seat belt patch curve may correspond to any suitable patch or region of an image (e.g., input image 102) that may be indicative of one or more seat belts, and the corresponding localization of a structured edge from the seat belt patch curve may indicate any suitable edge or feature based at least in part on the seat belt patch curve.

[0034] 5, the local predictor 104 may or may not select the two highest points in a particular curve, since the selection may be inaccurate since the two highest points in the particular curve may be located near the highest peak. Referring to FIG. 5, if the curve cuts into two parts, the two peaks may be separated and both peaks may be obtained by the local predictor 104 by finding the maximum point across each point. A one-dimensional two-class classification task may be performed using idx optimalThe inter-class distance can be defined by the following formula: optimal However, any variation thereof may be used:

number

number

number

number

number

number

[0035] The local predictor 104 uses idx optimal The local predictor 104 can solve the nonlinear optimization task by determining at least [1, idx optimal ] and (idx optimal ,L] to find the location of the largest element in left and idx right The function values ​​of each of these can be calculated as peak left and peak right It could be.

[0036] The seatbelt pixel intensity can be different and change in response to various ambient lighting. Ambient lighting can include various light sources such as the vehicle's interior lights. The local predictor 104 can learn the intensity distribution of the pixels. δ min and δ max can be set as the lower and upper bounds for the seat belt localization case. patch The weighting strength of (x, y, θ) is d intensity and may be calculated via the following formula, although any variation thereof may be used:

number

number

[0037] In one embodiment, f intensity is defined by the following formula, although any variation thereof may be used:

number

[0038] In various embodiments, the seat belt surface is smooth with minimal intensity variation. Intensity variance can be used within the region of interest ω as a criterion for assessing smoothness. In one embodiment, d smoothness is defined by the following, although any variation thereof may be used:

number

number

[0039] The local predictor 104 may output pixel candidates 106, which may include one or more indications of one or more pixels of the input image 102 that may potentially correspond to a seat belt shown in the input image 102. The pixel candidates 106 may be input to a global assembler 108, which may remove inaccurate candidates from the pixel candidates 106 and selectively assemble qualified seat belt pixel candidates.

[0040] The global assembler 108 may be a collection of one or more hardware and / or software computational resources having executable instructions that, when executed, compiles positive seat belt portion candidates by eliminating false positives based on seat belt attributes such as shape and position. The global assembler 108 may apply a set of constraints to the pixel candidates 106 to narrow down the pixel candidates 106 to obtain qualified pixel candidates 114. The set of constraints may include features such as parameters of the camera that captured the input image 102 (e.g., camera lens dimensions, focal point, principal point, distortion parameters), standard seat belt size range (e.g., width and / or length range), standard seat belt features and / or parameters (e.g., standard seat belt color, material), vehicle parameters (e.g., vehicle layout, vehicle components), etc. The global assembler 108 may apply a set of constraints to the pixel candidates 106 such that pixel candidates in the pixel candidates 106 that do not adhere to or otherwise comply with the set of constraints may be eliminated to determine eligible pixel candidates 114 (e.g., pixel candidates that represent seat belts having widths significantly longer than the standard seat belt width range may be eliminated).

[0041] Global assembler 108 may include automatic position mask generation 110 and width range estimation 112. Automatic position mask generation 110 may be a collection of one or more hardware and / or software computational resources having executable instructions to generate one or more image masks of an image, such as input image 102, showing one or more seat belts. Width range estimation 112 may be a collection of one or more hardware and / or software computational resources having executable instructions to estimate a range of widths of seat belts shown in an image, e.g., input image 102.

[0042] In one embodiment, for a three-point seat belt, the seat belt has three anchors, called points: tr The top right anchor, denoted by A bl The lower left anchor indicated by br The lower right anchor is shown by. The belts coming from the upper right anchor and the lower right anchor can be inserted into the lower left anchor using a buckle. The techniques described herein relating to the upper right anchor and the lower right anchor can be similarly applied to the lower right anchor, the lower left anchor, or the anchors of any suitable restraining device. In some embodiments, the seat belt anchors are movable. That is, for example, one or more seat belt anchors of a three-point seat belt can be adjusted, rotated, and / or titled in various directions. In some embodiments, one or more seat belt anchors are adjusted via tilt to a particular angle. In some cases, the seat belt anchors are connected to a mounting plate attached to a side door of the passenger vehicle. The seat belt anchors can be moved relative to the mounting plate (e.g., by sliding up, down, left, or right).

[0043] There may be a constant force inside the anchor to tighten the belt using a spring, and the location distribution of the seat belt pixels observed through a fixed camera may be within a limited area. In some instances, the global assembler 108 includes various machine learning algorithms to learn the location distribution of the seat belt pixels. In various embodiments, the seat belt does not generally appear on the steering wheel or other in a vehicle system (e.g., entertainment system, driver assistance system). In some instances, a range of seat belt widths in an image frame is estimated by the global assembler 108 to filter seat belt candidates. The global assembler 108 may process multiple seat belts in a frame. In some embodiments, one or more processes of the global assembler 108 are executed in parallel using algorithm level parallelism. Algorithm level parallelism may refer to a process in which processes that can be executed in parallel (e.g., processes or operations executed by a system for seat belt localization) are first identified and then executed in parallel. The global assembler 108 may perform various optimization techniques, such as image downsampling, pixel step adjustment, angle sampling, and the like.

[0044] The automatic position mask generation 110 can generate a seat belt position mask, which may be a visible indication of the position of a seat belt in an image, e.g., the input image 102. The automatic position mask generation 110 can include one or more software programs that can analyze one or more aspects of a camera, e.g., a camera that captured the input image 102. The automatic position mask generation 110 can analyze various configurations of the camera to generate a seat belt position mask, e.g., camera calibration, camera pose, camera lens type, etc. The automatic position mask generation 110 can include one or more components that can perform camera calibration, camera localization, and 3D reconstruction. The camera calibration can include one or more functions and / or processes that can calibrate the camera (e.g., infrared (IR) camera, RGB (red-green-blue) camera, color camera, etc.) and process the image from the camera by processing the internal attributes of the camera, e.g., focal lens, principal point, or undistortion parameters. Camera localization may include one or more functions and / or processes that can provide six degrees of freedom (6DOF) position information of the camera in a given world coordinate system. 3D reconstruction may include one or more functions and / or processes that can determine the coordinates of any point in 3D space through the use of encoded tags. Camera localization may include processes that can determine the 6DOF pose of the camera with respect to the vehicle coordinate system and obtain the coordinates or location of the seat belt anchor within the camera coordinate system. Camera localization may determine the coordinates or location of the seat belt anchor using one or more models of the vehicle. 3D reconstruction may include processes that can calculate the relative relationship between the seat belt anchor and the camera coordinate system.

[0045] In one embodiment, a camera (e.g., the camera capturing the input image 102) is calibrated with intrinsic parameters K and extrinsic parameters R and T. The seat belt is tr (X tr ,Y tr ,Z tr ) and A bl (X bl ,Y bl ,Z bl The automatic location mask generation 110 can obtain the corresponding coordinates of the anchor, referred to as the anchor's location, in the image coordinate system. In one embodiment, the image location of the anchor is represented by A i (x i ,y i ), i=tr, bl, where any of the following formulas or variations thereof may be used:

number

[0046] The majority of the seat belt position distribution can fall into an ellipse using the two anchors as the ends of the long axis. In one embodiment, the distance of the long axis is d major and may be defined by the following formula, although any variation thereof may be used:

number

[0047] The automatic position mask generation 110 can obtain the short axis distance through one or more machine learning algorithms that can analyze various seat belt fastening processes. In one embodiment, the short axis distance is d minor and the set position mask is denoted by S location and is defined by the following formula, although any variation thereof may be used:

number

number

number

number

[0048] The width range estimation 112 may estimate the width range, also referred to as width size, of the seat belt in an image (e.g., the input image 102). A seat belt standard width may be in the range of 46-49 mm (e.g., 2 inches), where most belts may generally be about 47 mm, or any suitable value in any suitable range. The width of the seat belt in the image may depend on the pose of the seat belt. If the belt is parallel to the camera imaging sensor or stretched towards the camera, the seat belt width may increase. If the belt surface is perpendicular to the image sensor or further away from the camera, the seat belt width may decrease. A lower limit τ of the seat belt width minmay be set based at least in part on one or more sensitivity requirements. min is set to τ to avoid including noise patterns. max is set to half of the

[0049] In one embodiment, the seat belt point in the middle of the two anchors is P s (X s ,Y s ,Z s ) and the standard seat belt width is d std where S pair The possible corresponding points of the other ends, denoted by

number

[0050] The arm distance of a normal passenger is d arm where the maximum width τ max In one embodiment, the new seat belt point coordinate set is S new and is defined by the following formula, although any variation thereof may be used:

number

[0051] The width range estimation 112 extracts the envelope of the new point locations and estimates the width upper bound τ max The seat belt width can be used to determine the structured edge threshold, τ max The global assembler 108 may be used to determine a patch size L that may be needed to be larger than the global assembler 108. The global assembler 108 may be used to determine the locations of multiple seat belts in a single image. The global assembler 108 may generate a location mask for each seat belt region.

[0052] The global assembler 108 can process an image (e.g., the input image 102) using one or more graphics processing units (GPUs). The global assembler 108 can process multiple patches simultaneously using multiple threads of one or more GPUs. The global assembler 108 can process seat belt regions in parallel using work streams of one or more GPUs. The global assembler 108 can process one or more areas of an image using one or more threads of one or more GPUs in parallel or simultaneously. The global assembler 108 can assign processing of multiple frames to different computational units of one or more GPUs to achieve frame-level acceleration. The global assembler 108 can downsample the input image to reduce the use of computational resources in processing the input image. The global assembler 108 can process an image (e.g., the input image 102) in multiple different directions in addition to other ways or pixel-by-pixel traversals of the image. The global assembler 108 can adjust the discretization of patch orientation angles to reduce computational resource usage. The global assembler 108 can assemble or otherwise determine eligible pixel candidates 114, which may include pixels of the input image 102 that correspond to one or more seat belts depicted in the input image 102.

[0053] The qualified pixel candidates 114 may include a set of seat belt pixel candidates for each seat belt region of interest. The qualified pixel candidates 114 may also be referred to as candidate pixels, pixel candidates, candidates, and / or variations thereof. The shape modeler 116 may be a collection of one or more hardware and / or software computational resources having executable instructions that, when executed, build a geometric seat belt shape model based on the output from the global assembler 108 (e.g., the qualified pixel candidates 114) and map the model onto the input image. The shape modeler 116 may remove noise candidates from the qualified pixel candidates 114 and model the seat belt shape.

[0054] In one embodiment, the set of candidates from the local predictor 104 is S predictor and the position mask is denoted by S location where S predictor Shape modeling in S modeling The candidates for are defined by the following formula, although any variation thereof may be used: S modeling ={(x,y)|(x,y)∈S location &&f predictor (x,y)≧γ pre} (twenty two), So, γ pre can be a threshold for predicting the number of responses, which can be proportional to the total number of patch orientations.

[0055] The shape modeler 116 may model the shape of the seat belt via a high-order polynomial curve. A high-order polynomial curve may refer to a polynomial curve having a degree higher than 2. The shape modeler 116 may model the shape of the seat belt via a polynomial curve of any suitable degree. The degree of the polynomial curve may correspond to the complexity of the shape of the seat belt, with higher degrees corresponding to more complex shapes (e.g., shapes with many curves, oscillations, etc.). In some instances, a passenger may push the seat belt away while driving, which may result in various locations along the x-axis, multiple y-values, which may result in an inapplicable polynomial model. In one embodiment, the shape modeler 116 transforms the global image coordinates to the anchor-local coordinate system via the following equation, although any variation thereof may be used:

number

[0056] FIG. 6 illustrates an example candidate map 600 according to at least one embodiment. Example 600 may illustrate a candidate map after transformation from global image coordinates to an anchor local coordinate system. Note that FIG. 6 illustrates one example of potential candidates and curve fits to the candidates, and the candidates and fitted curves may be any variations thereof. The potential candidates may correspond to any suitable pixels or regions of pixels of an image (e.g., input image 102) that may be indicative of one or more seat belts, and the corresponding curve fits to the potential candidates may indicate any suitable seat belt shapes.

[0057] The example 600 may show seat belt candidate pixels and a curve (e.g., a high order polynomial curve) fit to the candidate pixels corresponding to the shape of the seat belt. Because different users may stretch their seat belts into different shapes, a high order polynomial regression to model the shape of the seat belt may be used. In one embodiment, the observed points are (x i ,y i ), where y i corresponds to the following formula, although any variation thereof may be used:

number

[0058] The shape modeler 116 can use various algorithms to reduce noise, for example, an M-estimator sample consensus (MSAC) algorithm. The shape modeler 116 can transform the curves to a global image plane for visualization. The shape modeler 116 can determine a model of seat belt shape from an image (e.g., the input image 102) showing a seat belt that at least approximates the shape and / or location of the seat belt. The shape modeler 116 can further determine whether a seat belt identified in an image (e.g., the input image 102) is correctly applied. In some instances, the shape modeler 116 provides the model of the seat belt to one or more neural networks trained to infer whether the seat belt is correctly applied based on the model. The one or more neural networks can determine whether the seat belt is correctly applied based on the seat belt model obtained from the shape modeler 116 and provide an indication of whether the seat belt is correctly applied to the shape modeler 116. The shape modeler 116 can visualize the modeled seat belt shape as well as whether the seat belt corresponding to the modeled seat belt shape is correctly applied as a modeled input image 118. In one embodiment, the modeled input image 118 is the input image 102 having a model of the visualized seat belt shape. The modeled input image 118 can show the input image 102 with the position and shape of the seat belt in the input image 102 indicated via one or more visualizations, such as the visual boundary of the seat belt. The modeled input image 118 can further include one or more visualizations that can indicate whether the seat belt is worn and / or correctly applied.

[0059] For example, referring to FIG. 1 , the modeled input image 118 includes a visual boundary of the seat belt and an indication that the seat belt is fastened and correctly applied, indicated by “Seat Belt: On”. The indication that the seat belt is fastened and correctly applied may be displayed in various ways, e.g., “Seat Belt: On”, “Seat Belt: Applied”, etc. The indication that the seat belt is fastened but incorrectly applied may be displayed in various ways, e.g., “Seat Belt: Off”, “Seat Belt: Incorrect”, etc. The indication that the seat belt is not fastened may be displayed in various ways, e.g., “Seat Belt: Off”, “Seat Belt: Not Applied”, etc. A correctly applied or properly placed seat belt may be a seat belt that is in a locked position and worn diagonally across the occupant’s chest. An incorrectly applied or improperly placed seat belt may be a seat belt that is not correctly applied or not in the proper place.

[0060] The system for seat belt localization can provide a model of the seat belt and an indication of whether the seat belt is properly applied to one or more systems of the vehicle. In some instances, the system for seat belt localization executes on one or more computer servers, where the model of the seat belt and the indication of whether the seat belt is properly applied are provided to one or more systems of the vehicle over one or more communication networks. The system of the vehicle can perform various actions as a result of obtaining the model of the seat belt and the indication of whether the seat belt is properly applied. The system of the vehicle can provide a warning indication if the seat belt is not fastened or is fastened but incorrectly applied. The warning indication can be an audio indication, such as a warning sound, a visual indication, such as a warning light, a physical indication, such as a warning vibration, or the like. The system of the vehicle can cause one or more propulsion systems of the vehicle to cease or stop propulsion of the vehicle if the seat belt is not fastened or is fastened but incorrectly applied. The vehicle's system can provide indications via one or more networks to various systems (e.g., safety monitoring systems) when a seat belt is not fastened or is fastened but applied incorrectly.

[0061] The system for seat belt localization may be a passive system. The system may not require modifications to existing systems of the vehicle to detect the seat belt of an occupant of the vehicle from one or more images of the seat belt. The system may be capable of passive detection of the seat belt, which may refer to detection that does not require input from one or more sensors of the vehicle or modifications to one or more systems of the vehicle (e.g., the system may not require the seat belt of the vehicle to have an identifying mark). The system for seat belt localization may combine models determined from multiple images to determine a final seat belt model. In some instances, the system for seat belt localization localizes and models a set of seat belts from a set of images showing an occupant and the corresponding seat belt, where the system combines the models to determine a final seat belt model of the occupant's corresponding seat belt. The system for seat belt localization combines or otherwise augments the determined seat belt models together to determine a final seat belt model such that the reliability of the final seat belt model is increased. The system for seat belt localization can combine seat belt models by evaluating similarities and differences between the seat belt models to determine a final seat belt model. The system for seat belt localization can combine estimations made from one or more images to determine a final seat belt model, increasing confidence in the final seat belt model.

[0062] In some instances, the system for seat belt localization may be used in conjunction with one or more active seat belt detection systems. An active seat belt detection system may refer to one or more systems in a vehicle that are used to directly detect and localize the vehicle's seat belts. An active seat belt detection system may include the use of various sensors, such as pressure sensors, weight sensors, motion sensors, or other sensors, to detect and localize the seat belts and may require modifications to the vehicle's existing systems (e.g., an active seat belt detection system may require the vehicle's seat belts to have an identifying mark). The system for seat belt localization may localize the seat belts and verify that the localized seat belts have been correctly determined using input from the one or more active seat belt detection systems. Alternatively, the system may receive one or more models of the localized seat belts from the one or more active seat belt detection systems and localize the seat belts via one or more images to verify that the localized seat belt models have been correctly determined. The system may receive or obtain input from one or more vehicle systems, e.g., an active seat belt detection system, to localize the seat belt. For example, a system for seat belt localization may obtain data from vehicle sensors, e.g., pressure sensor data, weight sensor data, motion sensor data, etc., and use the data in accordance with the techniques described above to localize and model the seat belt from an image of the seat belt in the vehicle. The sensor data may be used to determine constraints for filtering pixel candidates (e.g., the sensor data may provide information indicative of the location or position of the seat belt, which may be used to eliminate poor pixel candidates that do not conform to the information provided by the sensor data).

[0063] The system for seat belt localization may localize and model the seat belt via a patch matching and patch refinement process, such as those mentioned above and / or various algorithms, such as a patch match algorithm, a mixed-resolution patch-matching (MRPM) algorithm, and / or variations thereof. The patch matching process may take image patches as input to one or more neural networks to extract patch characteristics and evaluate similarities of the patches. The system for seat belt localization may define or otherwise provide seat belt patch characteristics, features, motions, etc., construct estimated seat belt patches using various patch matching processes, and refine the estimated seat belt patches to determine a final seat belt model. Patch matching may be on spatial and / or temporal dimensions.

[0064] FIG. 7 illustrates an example 700 of multiple seat belt localization according to at least one embodiment. An input image 702, a seat belt localizer 704, and a modeled input image 706 may be in accordance with those described with respect to FIGS. 1-6. The seat belt localizer 704 may be a collection of one or more hardware and / or software computational resources having executable instructions that, when executed, model one or more restraining devices from one or more images. The seat belt localizer 704 may acquire or otherwise receive an input image 702, perform an initial classification of areas of the input image 702 that represent restraining devices (e.g., seat belts), apply a set of constraints to areas of the input image 702 to refine the initial classification to obtain a refined classification, and generate a model of the restraining device based at least in part on the refined classification. The seat belt localizer 704 may identify and model multiple seat belts depicted in an image (e.g., the input image 702).

[0065] 7, the input image 702 may be an image captured from one or more cameras inside a vehicle. The input image 702 may show the occupants of the vehicle, including a passenger, which may be shown on the left, and a driver, which may be shown on the right. The input image 702 may show the passenger correctly fastening the seat belt and the driver correctly fastening the seat belt. The seat belt localizer 704 may analyze the input image 702 to identify and model the seat belt in the input image 702. The seat belt localizer 704 may visualize the modeled seat belt via the modeled input image 706.

[0066] 7, a modeled input image 706 may show an input image 702 with a seat belt and an indicated seat belt status (e.g., orientation / location). The seat belt localizer 704 may perform an initial classification for a passenger shown in the input image 702, refine the initial classification to obtain a refined classification, and generate a model based on the refined classification of the passenger's seat belt, and perform a second initial classification for a driver shown in the input image 702, refine the second initial classification to obtain a second refined classification, and generate a second model based on the second refined classification of the driver's seat belt.

[0067] The modeled input image 706 may include a first visible indication of a passenger's seat belt with a corresponding label indicating the status of the passenger's seat belt. With reference to FIG. 7, the modeled input image 706 may show, on the left side, a box indicating the passenger's seat belt and the label "Seat Belt: On" indicating that the passenger's seat belt is fastened and correctly applied. The modeled input image 706 may include a second visible indication of a driver's seat belt with a corresponding label indicating the status of the driver's seat belt. With reference to FIG. 7, the modeled input image 706 may show, on the right side, a box indicating the driver's seat belt and the label "Seat Belt: On" indicating that the driver's seat belt is fastened and correctly applied.

[0068] FIG. 8 illustrates an example 800 of seat belt localization according to at least one embodiment. An input image 802, a seat belt localizer 804, and a modeled input image 806 may be in accordance with those described with respect to FIGS. 1-6. The seat belt localizer 804 may be a collection of one or more hardware and / or software computational resources having executable instructions that, when executed, model one or more restraining devices from one or more images. The seat belt localizer 804 may acquire or otherwise receive an input image 802, perform an initial classification of areas of the input image 802 that represent restraining devices (e.g., seat belts), apply a set of constraints to areas of the input image 802 to refine the initial classification to obtain a refined classification, and generate a model of the restraining device based at least in part on the refined classification. The seat belt localizer 804 may identify and model a seat belt depicted in an image (e.g., the input image 802).

[0069] 8, input image 802 may be an image captured from one or more cameras inside a vehicle. Input image 802 may show passengers in the vehicle, including the driver, who may be shown in the center. Input image 802 may show that the driver is wearing a seat belt incorrectly (e.g., the driver has the seat belt behind their back). Seat belt localizer 804 may analyze input image 802 to identify and model the seat belt in input image 802. Seat belt localizer 804 may visualize the modeled seat belt via modeled input image 806.

[0070] With reference to Figure 8, modeled input image 806 may show input image 802 with a seat belt and an indicated seat belt status (e.g., orientation / location). Modeled input image 806 may include a first visible indication of a driver's seat belt with a corresponding label indicating the driver's seat belt status. With reference to Figure 8, modeled input image 806 may show a box indicating the driver's seat belt and the label "Seat belt: off" indicating that the driver's seat belt is incorrectly fastened.

[0071] FIG. 9 illustrates an example 900 of seat belt localization according to at least one embodiment. The input image 902, seat belt localizer 904, and modeled input image 906 may be in accordance with those described with respect to FIGS. 1-6. The seat belt localizer 904 may be a collection of one or more hardware and / or software computational resources having executable instructions that, when executed, model one or more restraining devices from one or more images. The seat belt localizer 904 may acquire or otherwise receive the input image 902, perform an initial classification of areas of the input image 902 that represent restraining devices (e.g., seat belts), apply a set of constraints to the areas of the input image 902 to refine the initial classification to obtain a refined classification, and generate a model of the restraining device based at least in part on the refined classification. The seat belt localizer 904 may identify and model a seat belt depicted in an image (e.g., input image 902).

[0072] 9, input image 902 may be an image captured from one or more cameras inside a vehicle. Input image 902 may show passengers in the vehicle, including the driver, who may be shown in the center. Input image 902 may show that the driver is wearing a seat belt incorrectly (e.g., the driver has the seat belt under their arm). Seat belt localizer 904 may analyze input image 902 to identify and model the seat belt in input image 902. Seat belt localizer 904 may visualize the modeled seat belt via modeled input image 906.

[0073] 9, modeled input image 906 may show input image 902 with a seat belt and an indicated seat belt status (e.g., orientation / location). Modeled input image 906 may include a first visible indication of a driver's seat belt with a corresponding label indicating the driver's seat belt status. With reference to FIG. 9, modeled input image 906 may show a box indicating the driver's seat belt and the label "Seat belt: off" indicating that the driver's seat belt is incorrectly fastened.

[0074] FIG. 10 illustrates an example 1000 of seat belt localization according to at least one embodiment. The modeled input image 1002 may be according to those described in accordance with FIGS. 1-6. The seat belt localizer may be a collection of one or more hardware and / or software computational resources having executable instructions that, when executed, model one or more restraining devices from one or more images. The seat belt localizer may acquire or otherwise receive an input image, perform an initial classification of an area of ​​the input image that represents a restraining device (e.g., a seat belt), apply a set of constraints to the area of ​​the input image to refine the initial classification to obtain a refined classification, generate a model of the restraining device based at least in part on the refined classification, and visualize the model as the modeled input image 1002.

[0075] The modeled input image 1002 may be based on an input image that may be captured from one or more cameras. The input image may be captured from a camera with a standard FOV. The seat belt localizer may identify and model one or more seat belts from an image of any FOV. The seat belt localizer may identify and model one or more seat belts from an image that may show portions of a seat belt. With reference to FIG. 10, the modeled input image 1002 may show a standard FOV image with a seat belt identified by a shaded region and a corresponding label "Seat Belt: On" indicating that the seat belt is fastened and correctly applied.

[0076] FIG. 11 illustrates an example 1100 of seat belt localization according to at least one embodiment. The modeled input image 1102 may be according to those described with respect to FIGS. 1-6. The seat belt localizer may be a collection of one or more hardware and / or software computational resources having executable instructions that, when executed, model one or more restraining devices from one or more images. The seat belt localizer may acquire or otherwise receive an input image, perform an initial classification of an area of ​​the input image that represents a restraining device (e.g., a seat belt), apply a set of constraints to the area of ​​the input image to refine the initial classification to obtain a refined classification, generate a model of the restraining device based at least in part on the refined classification, and visualize the model as the modeled input image 1102.

[0077] The modeled input image 1102 may be based on an input image, which may be captured from one or more cameras. The input image may show a driver pushing a seat belt away from the driver's body. The seat belt localizer may identify and model one or more seat belts from an image showing one or more seat belts in any suitable location (e.g., pushed aside, extended, retracted). With reference to FIG. 11, the modeled input image 1102 may show an image with a seat belt identified by a dotted curve and a corresponding label "seat belt: off" indicating that the seat belt is worn but not applied correctly.

[0078] It should be noted that although FIGS. 7-11 illustrate examples of a seat belt localizer that identifies, models, visualizes, and labels one or more seat belts, the seat belt localizer may identify, model, visualize, and label one or more seat belts in any suitable manner. The seat belt localizer may visualize one or more seat belts in an image by indicating one or more seat belts via an outline, a box, a curve, a bounding box, a shaded area, a patterned area, or other indication. The seat belt localizer may label one or more seat belts in an image via a variety of indications, such as a symbol, a letter, a label, or other indication. An indication that a seat belt is fastened and correctly applied may be indicated in a variety of ways, such as by displaying text on a display screen within the vehicle indicating: "Seat belt: on," "Seat belt: applied," "Seat belt: correctly applied," etc. An indication that a seat belt is incorrectly fastened may be displayed in a variety of ways, e.g., "Seat belt: off", "Seat belt: incorrect", "Seat belt: incorrectly applied", etc. An indication that a seat belt is not fastened may be shown in a variety of ways, e.g., "Seat belt: off", "Seat belt: not applied", "Seat belt not applied", etc. The indication may also include an audio indication (e.g., a computer voice indication), a visual indication (e.g., an indicator light, a switch), a physical indication (e.g., an indicator vibration), etc.

[0079] Referring now to FIG. 12, each block of the method 1200 described herein includes computational processes that may be performed using any combination of hardware, firmware, and / or software. For example, various functions may be performed by a processor executing instructions stored in a memory. The method may also be implemented as computer usable instructions stored on a computer storage medium. The method may be provided by a standalone application, a service, or a hosted service (standalone or in combination with another hosted service), or a plug-in to another product, to name a few. Additionally, the method 1200 is described with respect to the system for seat belt localization of FIG. 1, by way of illustration. However, these methods may additionally or alternatively be performed by any one system or any combination of systems, including but not limited to those described herein.

[0080] 12 is a flow diagram illustrating a method 1200 for generating a model of an input image, according to some embodiments of the present disclosure. The method 1200 includes, at block 1202, acquiring an input image. The input image may represent a subject, such as a driver, a passenger, or other occupant, and a restraining device applied to the subject, such as a seat belt, safety harness, or other restraining device. The input image may be an image captured from one or more image capture devices, such as a camera or other device, that may be located inside the vehicle.

[0081] The method 1200 includes determining pixel candidates at block 1204. A system performing at least a portion of the method 1200 may perform an initial classification of areas of the image as representing restraint devices. The initial classification may result in the determination of pixel candidates. For any given pixel of the input image, the system may use the pixel and its neighboring pixel information to predict whether the pixel belongs to a seat belt or is a background pixel. The system may determine areas of the input image that show or otherwise represent restraint devices (e.g., seat belts) to determine pixel candidates.

[0082] The method 1200 includes, at block 1206, filtering out false positives and determining eligible pixel candidates. A system performing at least a portion of the method 1200 may apply a set of constraints to areas of the image to refine the initial classification to obtain a refined classification. The refined classification may result in the determination of eligible pixel candidates. The set of constraints may include features such as parameters of the camera that captured the input image (e.g., camera lens dimensions, focal point, principal point, distortion parameters), standard seat belt size ranges (e.g., width and / or length ranges), standard seat belt features and / or parameters (e.g., standard seat belt colors, materials), vehicle parameters (e.g., vehicle layout, vehicle components), etc. The system may apply the set of constraints to the pixel candidates such that pixels of the pixel candidates that do not adhere to or otherwise comply with the set of constraints may be filtered out to determine eligible pixel candidates.

[0083] The method 1200 includes, at block 1208, constructing a model based on the qualified pixel candidates. A system performing at least a portion of the method 1200 can generate a model of the restraint device based at least in part on the refined classification. The system can generate the model based on pixels of the qualified pixel candidates that correspond to the seat belt. The method 1200 includes, at block 1210, mapping the model to the input image. The system can visualize the model on the input image as a modeled input image. The modeled input image can include the location / orientation / position of the seat belt as well as an indication of the seat belt status indicating whether the seat belt is fastened, not fastened, applied, or incorrectly applied.

[0084] Exemplary Autonomous Vehicle 13A illustrates an example autonomous vehicle 1300 according to some embodiments of the present disclosure. Autonomous vehicle 1300 (alternatively referred to herein as “vehicle 1300”) may include, but is not limited to, a passenger vehicle, such as a car, a truck, a bus, a first responder vehicle, a shuttle, an electric or moped, a motorcycle, a fire engine, a police vehicle, an ambulance, a boat, a construction vehicle, a submarine, a drone, and / or another type of vehicle (e.g., unmanned and / or carrying one or more passengers). Autonomous vehicles are generally described in terms of automation levels as 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). The mobile vehicle 1300 may be capable of functionality according to one or more of the autonomous driving levels Levels 3 through 5. For example, the mobile vehicle 1300 may be capable of conditional automation (Level 3), high automation (Level 4), and / or full automation (Level 5), depending on the embodiment.

[0085] The mobile vehicle 1300 may include components such as a chassis, body, wheels (e.g., 2, 4, 6, 8, 18, etc.), tires, axles, and other components of the mobile vehicle. The mobile vehicle 1300 may include a propulsion system 1350, such as an internal combustion engine, a hybrid power plant, a fully electric engine, and / or another propulsion system type. The propulsion system 1350 may be connected to a drive train of the mobile vehicle 1300, which may include a transmission, to enable propulsion of the mobile vehicle 1300. The propulsion system 1350 may be controlled in response to receiving a signal from a throttle / accelerator 1352.

[0086] A steering system 1354, which may include a steering wheel, may be used to steer the vehicle 1300 (e.g., along a desired course or route) when the propulsion system 1350 is operating (e.g., when the vehicle is moving). The steering system 1354 may receive signals from a steering actuator 1356. The steering wheel may be an option for fully automated (Level 5) functionality.

[0087] Brake sensor system 1346 may be used to operate vehicle brakes in response to receiving a signal from a brake actuator 1348 and / or a brake sensor.

[0088] A controller 1336, which may include one or more CPUs, system on chip (SoC) 1304 (FIG. 13C), and / or GPUs, can provide signals (e.g., representations of commands) to one or more components and / or systems of the vehicle 1300. For example, the controller can send signals to operate vehicle brakes via one or more brake actuators 1348, operate a steering system 1354 via one or more steering actuators 1356, and / or operate a propulsion system 1350 via one or more throttle / accelerators 1352. The controller 1336 can include one or more on-board (e.g., integrated) computing devices (e.g., supercomputers) that process sensor signals and output operational commands (e.g., signals representing commands) to enable autonomous driving and / or to assist a driver in driving the vehicle 1300. The controllers 1336 may include a first controller 1336 for autonomous driving functions, a second controller 1336 for functional safety functions, a third controller 1336 for artificial intelligence functions (e.g., computer vision), a fourth controller 1336 for infotainment functions, a fifth controller 1336 for redundancy in emergency situations, and / or other controllers. In some instances, a single controller 1336 may handle two or more of the aforementioned functions, and two or more controllers 1336 may handle a single function, and / or any combination thereof.

[0089] Controller 1336 may provide signals to control one or more components and / or systems of vehicle 1300 in response to sensor data (e.g., sensor inputs) received from one or more sensors. The sensor data may be received from, for example, and without limitation, global navigation satellite system sensors 1358 (e.g., global positioning system sensors), RADAR sensors 1360, ultrasonic sensors 1362, LIDAR sensors 1364, inertial measurement unit (IMU) sensors 1366 (e.g., accelerometers, gyroscopes, magnetic compasses, magnetometers, etc.), microphones 1396, stereo cameras 1368, wide-view cameras 1370 (e.g., fisheye cameras), infrared cameras 1372, surround cameras 1374 (e.g., 360 degree cameras), long-range and / or medium-range cameras 1398, speed sensors 1344 (e.g., for measuring the speed of the moving vehicle 1300), vibration sensors 1342, steering sensors 1340, brake sensors 1346 (e.g., as part of a brake sensor system 1346), and / or other sensor types.

[0090] One or more of the controllers 1336 may receive input (e.g., represented by input data) from the instrument cluster 1332 of the vehicle 1300 and provide output (e.g., represented by output data, display data, etc.) via a human-machine interface (HMI) display 1334, an audible annunciator, a loudspeaker, and / or other components of the vehicle 1300. The output may include information such as vehicle velocity, speed, time, map data (e.g., HD map 1322 of FIG. 13C), position data (e.g., the position of the vehicle 1300 on a map, etc.), direction, positions of other vehicles (e.g., occupancy grid), information about objects and object situations as perceived by the controller 1336, etc. For example, the HMI display 1334 may display information regarding the presence of one or more objects (e.g., road signs, warning signs, changing traffic signals, etc.) and / or a driving maneuver that the moving vehicle has performed, is performing, or will perform (e.g., changing lanes now, taking exit 34B in 2 miles, etc.).

[0091] The mobile vehicle 1300 further includes a network interface 1324 capable of communicating over one or more networks using one or more wireless antennas 1326 and / or a modem. For example, the network interface 1324 may be capable of communicating over LTE, WCDMA, UMTS, GSM, CDMA2000, etc. The wireless antennas 1326 may also enable communication between objects in the environment (e.g., mobile vehicles, mobile devices, etc.) using local area networks such as Bluetooth, Bluetooth LE, Z-Wave, ZigBee, etc., and / or low power wide-area networks (LPWANs) such as LoRaWAN, SigFox, etc.

[0092] 13B is an illustration of camera positions and fields of view of the example autonomous vehicle 1300 of FIG. 13A, according to some embodiments of the present disclosure. The cameras and their respective fields of view are one example embodiment and are not intended to be limiting. For example, additional and / or alternative cameras may be included and / or the cameras may be located at different positions on the vehicle 1300.

[0093] The camera type of the camera may include, but is not limited to, a digital camera that may be adapted for use with the components and / or systems of the mobile vehicle 1300. The camera may operate at automotive safety integrity level (ASIL) B and / or at another ASIL. The camera type may be capable of any image capture rate, such as 60 frames per second (fps), 120 fps, 240 fps, etc., depending on the embodiment. The camera may be capable of using a rolling shutter, a global shutter, another type of shutter, or a combination thereof. In some instances, the color filter array may include a red clear clear clear (RCCC) color filter array, a red clear clear blue (RCCB) color filter array, a red blue green clear (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, clear pixel cameras, such as cameras with RCCC, RCCB, and / or RBGC color filter arrays, may be used in an effort to increase light sensitivity.

[0094] In some instances, one or more of the cameras may 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 mono camera may be installed to provide functions including lane departure warning, traffic sign assist, and intelligent headlamp control. One or more of the cameras (e.g., all of the cameras) may simultaneously record and provide image data (e.g., video).

[0095] One or more of the cameras may be mounted in a mounting part, such as a custom designed (3D printed) part, to filter out stray light and reflections from within the vehicle (e.g., reflections from the dashboard reflected in the windshield mirror) that may interfere with the camera's image data capture capabilities. With reference to a side mirror mounting part, the side mirror part may be custom 3D printed such that the camera mounting plate fits the shape of the side mirror. In some instances, the camera may be integrated into the side mirror. For side view cameras, the cameras may also be integrated into four posts at each corner of the cabin.

[0096] A camera (e.g., a forward-facing camera) with a field of view that includes a portion of the environment in front of the vehicle 1300 may be used for surround view to help identify the forward path and obstacles and, with the aid of one or more controllers 1336 and / or control SoCs, provide information essential to generating an occupancy grid and / or determining a preferred vehicle path. Forward-facing cameras may be used to perform many of the same ADAS functions as LIDAR, including emergency braking, pedestrian detection, and collision avoidance. Forward-facing cameras may also be used for ADAS functions and systems, including other functions such as lane departure warning ("LDW"), autonomous cruise control ("ACC"), and / or traffic sign recognition.

[0097] A variety of cameras may be used in a forward-facing configuration, including, for example, a monocular camera platform including a complementary metal oxide semiconductor (CMOS) color imager. Another example may be a wide-view camera 1370 that may be used to capture objects entering the view from the periphery (e.g., pedestrians, crossing traffic, or bicycles). Although only one wide-view camera is shown in FIG. 13B, there may be any number of wide-view cameras 1370 in the vehicle 1300. In addition, a long-range camera 1398 (e.g., a long-view stereo camera pair) may be used for depth-based object detection, especially for objects for which the neural network has not yet been trained. The long-range camera 1398 may also be used for object detection and classification, as well as basic object tracking.

[0098] One or more stereo cameras 1368 may also be included in the forward-facing configuration. The stereo camera 1368 may include an integrated control unit with an extensible processing unit that may provide programmable logic (e.g., FPGA) and a multi-core microprocessor with a CAN or Ethernet interface integrated on a single chip. Such a unit may be used to generate a 3D map of the vehicle's environment, including distance estimates for all points in the image. An alternative stereo camera 1368 may include a compact stereo vision sensor that may include two camera lenses (one left and one right) and an image processing chip that can measure distances from the vehicle to objects of interest and use the generated information (e.g., metadata) to activate autonomous emergency braking and lane departure warning functions. Other types of stereo cameras 1368 may be used in addition to or instead of those described herein.

[0099] A camera having a field of view that includes a portion of the environment to the side of the moving vehicle 1300 (e.g., a side-view camera) may be used for surround view, providing information used to create and update the occupancy grid and to generate side impact collision warnings. For example, surround cameras 1374 (e.g., four surround cameras 1374 as shown in FIG. 13B) may be positioned around the moving vehicle 1300. The surround cameras 1374 may include wide-view cameras 1370, fish-eye cameras, 360-degree cameras, and / or the like. For example, four fish-eye cameras may be positioned at the front, back, and sides of the moving vehicle. In an alternative arrangement, the moving vehicle may use three surround cameras 1374 (e.g., left, right, and rear) and may utilize one or more other cameras (e.g., a forward-facing camera) as a fourth surround view camera.

[0100] A camera having a field of view that includes the portion of the environment behind the moving vehicle 1300 (e.g., a rear-view camera) may be used for parking assistance, surround view, rear collision warning, and creating and updating an occupancy grid. As described herein, a wide variety of cameras may be used, including, but not limited to, cameras that are also suitable as forward-facing cameras (e.g., long-range and / or mid-range cameras 1398, stereo cameras 1368, infrared cameras 1372, etc.).

[0101] FIG. 13C is a block diagram of an example system architecture of the example autonomous vehicle 1300 of FIG. 13A, according to some embodiments of the present disclosure. It should be understood that this and other arrangements described herein are merely illustrative. Other arrangements and elements (e.g., machines, interfaces, functions, sequences, groupings of functions, etc.) may be used in addition to or instead of those shown, and some elements may be omitted altogether. Furthermore, many of the elements described herein are functional entities that may be implemented as separate or distributed components or in conjunction with other components, and in any suitable combination and location. Various functions described herein as being performed by an entity may be implemented by hardware, firmware, and / or software. For example, various functions may be implemented by a processor executing instructions stored in a memory.

[0102] Each of the components, features, and systems of the mobile vehicle 1300 of FIG. 13C are shown as connected via a bus 1302. The bus 1302 may include a controller area network (CAN) data interface (alternatively referred to herein as a “CAN bus”). The CAN may be a network within the mobile vehicle 1300 that is used to help control various features and functions of the mobile vehicle 1300, such as 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 may be read to find steering angle, ground speed, engine revolutions per minute (RPM), button positions, and / or other vehicle status indicators. The CAN bus may be ASIL B compliant.

[0103] The bus 1302 is described herein as being a CAN bus, but this is not intended to be limiting. For example, FlexRay and / or Ethernet may be used in addition to or as an alternative to a CAN bus. Additionally, a single line is used to represent the bus 1302, but this is not intended to be limiting. There may be any number of buses 1302, which may include, for example, 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 instances, two or more buses 1302 may be used to perform different functions and / or may be used for redundancy. For example, a first bus 1302 may be used for collision avoidance functions and a second bus 1302 may be used for actuation control. In any instance, each bus 1302 may communicate with any of the components of the mobile vehicle 1300, and two or more buses 1302 may communicate with the same components. In some examples, each SoC 1304, each controller 1336, and / or each computer in the vehicle may have access to the same input data (e.g., input from sensors in the vehicle 1300) and may be connected to a common bus, such as a CAN bus.

[0104] The mobile vehicle 1300 may include one or more controllers 1336, such as those described herein with respect to FIGURE 13A. The controller 1336 may be used for a variety of functions. The controller 1336 may be coupled to any of a variety of other components and systems of the mobile vehicle 1300 and may be used for control of the mobile vehicle 1300, artificial intelligence of the mobile vehicle 1300, infotainment for the mobile vehicle 1300, and / or the like.

[0105] The mobile vehicle 1300 may include a system on a chip (SoC) 1304. The SoC 1304 may include a CPU 1306, a GPU 1308, a processor 1310, a cache 1312, an accelerator 1314, a data store 1316, and / or other components and features not shown. The SoC 1304 may be used to control the mobile vehicle 1300 in a variety of platforms and systems. For example, the SoC 1304 may be coupled in a system (e.g., the system of the mobile vehicle 1300) having an HD map 1322 that can obtain map refreshes and / or updates via a network interface 1324 from one or more servers (e.g., server 1378 of FIG. 13D).

[0106] CPU 1306 may include a CPU cluster or CPU complex (alternatively referred to herein as a "CCPLEX"). CPU 1306 may include multiple cores and / or L2 caches. For example, in some embodiments, CPU 1306 may include eight cores in a coherent multiprocessor configuration. In some embodiments, CPU 1306 may include four dual-core clusters, each with its own dedicated L2 cache (e.g., a 2M B L2 cache). CPU 1306 (e.g., CCPLEX) may be configured to support concurrent cluster operation allowing any combination of clusters of CPUs 1306 to be active at any given time.

[0107] The CPU 1306 may implement power management capabilities including one or more of the following features: individual hardware blocks may be automatically clock gated when idle to conserve dynamic power, each core clock may be gated when the core is not actively executing instructions by execution of WFI / WFE instructions, each core may be independently power gated, each core cluster may be independently clock gated when all cores are clock gated or power gated, and / or each core cluster may be independently power gated when all cores are power gated. The CPU 1306 may further implement an enhanced algorithm for managing power states where allowed power states and expected wake-up times are specified and the hardware / microcode determines the best power state for entering the cores, clusters, and CCPLEXes. The processing cores may support simplified power state entry sequences in software where work is offloaded to the microcode.

[0108] The GPU 1308 may include an integrated GPU (alternatively referred to herein as an "iGPU"). The GPU 1308 may be programmable and efficient for parallel workloads. In some instances, the GPU 1308 may use an enhanced tensor instruction set. The GPU 1308 may include one or more streaming microprocessors, where each streaming microprocessor may include an L1 cache (e.g., an L1 cache having at least 96 KB storage capacity) and two or more of the streaming microprocessors may share a cache (e.g., an L2 cache having 512 KB storage capacity). In some embodiments, the GPU 1308 may include at least eight streaming microprocessors. The GPU 1308 may use a computer-based application programming interface (API). In addition, the GPU 1308 may use one or more parallel computing platforms and / or programming models (e.g., NVIDIA's CUDA).

[0109] The GPU 1308 may be power optimized for best performance in automotive and embedded use cases. For example, the GPU 1308 may be fabricated on FinFETs (Fin field-effect transistors). However, this is not intended to be limiting and the GPU 1308 may be fabricated using other semiconductor fabrication processes. Each streaming microprocessor may incorporate several mixed-precision processing cores partitioned into multiple blocks. For example, and without limitation, 64 PF32 cores and 32 PF64 cores may be partitioned into four processing blocks. In such an example, each processing block may be assigned 16 FP32 cores, 8 FP64 cores, 16 INT32 cores, two mixed-precision NVIDIA tensor cores for deep learning matrix operations, an L0 instruction cache, a warp scheduler, a dispatch unit, and / or a 64KB register file. In addition, the streaming microprocessor may include independent parallel integer and floating-point data paths to provide efficient execution of workloads having a mix of computational and addressing operations. Streaming microprocessors may include independent thread scheduling capabilities to allow finer grain synchronization and coordination among concurrent threads. Streaming microprocessors may include a combined L1 data cache and shared memory unit to improve performance while simplifying programming.

[0110] The GPU 1308 may include high bandwidth memory (HBM) and / or a 16GB HBM2 memory subsystem to provide about 900GB / sec peak memory bandwidth in some instances. In some instances, synchronous graphics random-access memory (SGRAM), such as graphics double data rate type five synchronous random-access memory (GDDR5), may be used in addition to or in place of the HBM memory.

[0111] The GPU 1308 may include unified memory technology including access counters to enable more accurate movement of memory pages to the processors that access them most frequently, thereby improving the efficiency of storage ranges shared between processors. In some instances, address translation service (ATS) support may be used to enable the GPU 1308 to directly access the CPU 1306 page tables. In such instances, when the GPU 1308 memory management unit (MMU) experiences a miss, an address translation request may be sent to the CPU 1306. In response, the CPU 1306 may consult its page table for a virtual-to-real mapping of the address and send the translation back to the GPU 1308. As such, the unified memory technology may enable a single unified virtual address space for both the CPU 1306 and GPU 1308 memories, thereby simplifying GPU 1308 programming and porting of applications to the GPU 1308.

[0112] In addition, the GPU 1308 may include access counters that can record the frequency of accesses of the GPU 1308 to the memory of other processors. The access counters can help ensure that memory pages are moved to the physical memory of the processor that is accessing the page most frequently.

[0113] The SoC 1304 may include any number of caches 1312, including those described herein. For example, the cache 1312 may include an L3 cache available to both the CPU 1306 and the GPU 1308 (e.g., connected to both the CPU 1306 and the GPU 1308). The cache 1312 may include a write-back cache that can record line state, such as by using a cache coherence protocol (e.g., MEI, MESI, MSI, etc.). The L3 cache may include 4 MB or more, depending on the embodiment, although smaller cache sizes may be used.

[0114] The SoC 1304 may include an arithmetic logic unit (ALU) that may be utilized in performing processing for any of the various tasks or operations (e.g., processing DNNs) of the vehicle 1300. In addition, the SoC 1304 may include a floating point unit (FPU) (or other math co-processor or numeric co-processor type) for performing mathematical operations within the system. For example, the SoC 1304 may include one or more FPUs integrated as execution units within the CPU 1306 and / or GPU 1308.

[0115] The SoC 1304 may include one or more accelerators 1314 (e.g., hardware accelerators, software accelerators, or a combination thereof). For example, the SoC 1304 may include a hardware acceleration cluster, which may include optimized hardware accelerators and / or large on-chip memory. The large on-chip memory (e.g., 4MB of SRAM) may enable the hardware acceleration cluster to accelerate neural networks and other operations. The hardware acceleration cluster may be used to complement the GPU 1308 and to offload some of the tasks of the GPU 1308 (e.g., to free up more cycles for the GPU 1308 to perform other tasks). As an example, the accelerator 1314 may be used for target workloads that are stable enough to be suitable for acceleration (e.g., perception, convolutional neural networks (CNNs), etc.). As used herein, the term "CNN" may include all types of CNNs, including region-based or regional convolutional neural networks (RCNNs) and Faster RCNNs (e.g., as used for object detection).

[0116] The accelerator 1314 (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 and optimized to perform image processing functions (e.g., CNN, RCNN, etc.). The DLA may further be optimized for a specific set of neural network types and floating point operations, as well as inference. The design of the DLA may provide more performance per millimeter than a general-purpose GPU, greatly exceeding the performance of a CPU. The TPU may perform several functions, including, for example, single instance convolution functions, supporting INT8, INT16, and FP16 data types for both features and weights, as well as post-processor functions.

[0117] The DLA can quickly and efficiently run neural networks, in particular CNNs, on processed or unprocessed data for any of a variety of functions, including but not limited to: CNNs for object identification and detection using data from camera sensors, CNNs for distance estimation using data from camera sensors, CNNs for emergency vehicle detection and identification using data from microphones, CNNs for face recognition and moving vehicle owner identification using data from camera sensors, and / or CNNs for security and / or safety related events.

[0118] The DLA can perform any function of the GPU 1308, and by using an inference accelerator, for example, a designer can target either the DLA or the GPU 1308 for any function. For example, a designer can focus on processing CNNs and floating point operations on the DLA and offload other functions to the GPU 1308 and / or other accelerators 1314.

[0119] The accelerator 1314 (e.g., 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 may provide a balance between performance and flexibility. For example, each PVA may include, but is not limited to, any number of reduced instruction set computer (RISC) cores, direct memory access (DMA), and / or any number of vector processors, for example.

[0120] The RISC cores may interact with an image sensor (e.g., an image sensor of any of the cameras described herein), an image signal processor, and / or the like. Each RISC core may include any amount of memory. The RISC cores may use any of a number of protocols, depending on the embodiment. In some instances, the RISC cores may run a real-time operating system (RTOS). The RISC cores may be implemented using one or more integrated circuit devices, application specific integrated circuits (ASICs), and / or memory devices. For example, the RISC cores may include an instruction cache and / or tightly coupled RAM.

[0121] The DMA may allow components of the PVA to access system memory independent of the CPU 1306. The DMA may support any number of features used to provide optimizations to the PVA, including but not limited to supporting multi-dimensional addressing and / or circular addressing. In some instances, the DMA may support up to six or more dimensions of addressing, which may include block width, block height, block depth, horizontal block stepping, vertical block stepping, and / or depth stepping.

[0122] The vector processor may be a programmable processor that may be designed to efficiently and flexibly execute computer vision algorithm programming and provide signal processing capabilities. In some instances, the PVA may include a PVA core and two vector processing subsystem partitions. The PVA core may include a processor subsystem, a DMA engine (e.g., two DMA engines), and / or other peripherals. The vector processing subsystem may act as the primary processing engine of the PVA and may include a vector processing unit (VPU), an instruction cache, and / or a vector memory (e.g., VMEM). The VPU core may include a digital signal processor, such as, for example, a single instruction, multiple data (SIMD), very long instruction word (VLIW) digital signal processor. The combination of SIMD and VLIW may increase throughput and speed.

[0123] Each vector processor may include an instruction cache and may be coupled to dedicated memory. As a result, in some instances, each vector processor may be configured to execute independently of other vector processors. In other instances, the vector processors included in a particular PVA may be configured to employ data parallelism. For example, in some embodiments, multiple vector processors included in a single PVA may execute the same computer vision algorithm, but on different regions of an image. In other instances, the vector processors included in a particular PVA may execute different computer vision algorithms simultaneously on the same image, or even execute different algorithms on sequential images or portions of an image. In particular, any number of PVAs may be included in a hardware accelerated cluster, and any number of vector processors may be included in each PVA. In addition, the PVAs may include additional error correcting code (ECC) memory to increase overall system security.

[0124] The accelerator 1314 (e.g., a hardware acceleration cluster) may include a computer vision network-on-chip and SRAM to provide high bandwidth, low latency SRAM for the accelerator 1314. In some instances, the on-chip memory may include, for example and without limitation, at least 4 MB of SRAM consisting of eight field configurable memory blocks that may be accessible by both the PVA and DLA. Each pair of memory blocks may include an advanced peripheral bus (APB) interface, configuration circuitry, controllers, and multiplexers. Any type of memory may be used. The PVA and DLA may access the memory through a backbone that provides the PVA and DLA with high speed access to the memory. The backbone may include a computer vision network-on-chip that interconnects the PVA and DLA to the memory (e.g., using the APB).

[0125] The computer vision network-on-chip may include an interface that determines, prior to transmission of any control signals / addresses / data, that both the PVA and DLA provide ready and valid signals. Such an interface may provide separate phases and separate channels for transmitting control signals / addresses / data, as well as burst type communication for continuous data transfer. This type of interface may follow ISO26262 or IEC61508 standards, although other standards and protocols may be used.

[0126] In some examples, the SoC 1304 may include a real-time ray tracing hardware accelerator, such as described in U.S. Patent Application No. 16 / 101,232, filed Aug. 10, 2018. The real-time ray tracing hardware accelerator may be used to quickly and efficiently determine the location and scale of objects (e.g., within a world model) to generate real-time visualization simulations for RADAR signal interpretation, for acoustic propagation synthesis and / or analysis, for simulation of SONAR systems, for general wave propagation simulation, for comparison against LIDAR data for localization and / or other functions, and / or for other uses. In some examples, one or more tree traversal units (TTUs) may be used to perform one or more ray tracing related operations.

[0127] The accelerator 1314 (e.g., a hardware accelerator cluster) has a variety of applications for autonomous driving. The PVA may be a programmable vision accelerator that can be used for critical processing stages in ADAS and autonomous vehicles. The capabilities of the PVA are well suited to the domain of algorithms that require predictable processing at low power and low latency. In other words, the PVA works well with semi-dense or dense regular computations on small data sets that require predictable execution times with low latency and low power. Hence, in the context of a platform for autonomous vehicles, the PVA is designed to run classical computer vision algorithms, since the PVA is efficient in object detection and operating on integer computations.

[0128] For example, according to one embodiment of the present technology, the PVA is used to perform computer stereo vision. A semi-global matching based algorithm may be used in some instances, but this is not intended to be limiting. Many applications for Level 3-5 autonomous driving require motion estimation / stereo matching on the fly (e.g., structure from motion, pedestrian recognition, lane detection, etc.). The PVA can perform computer stereo vision functions with input from two monocular cameras.

[0129] In some instances, the PVA may be used to perform dense optical flow. For example, the PVA may be used to process raw RADAR data (e.g., using a 4D Fast Fourier Transform) to provide a processed RADAR signal before firing the next RADAR pulse. In other instances, the PVA is used for time of flight depth processing, for example, by processing raw time of flight data to provide processed time of flight data.

[0130] DLA can be used to implement any type of network to enhance control and driving safety, including, for example, a neural network that outputs a measure of the confidence of each object detection. Such confidence value can be interpreted as a probability or as providing a relative "weight" of each detection compared to other detections. This confidence value allows the system to make further decisions regarding which detections should be considered true positive detections rather than false positive detections. For example, the system can set a confidence threshold and consider only detections above the threshold as true positive detections. In an automatic emergency braking (AEB) system, a false positive detection would cause a moving vehicle to automatically perform emergency braking, which is clearly undesirable. Therefore, only the most confident detections should be considered to trigger AEB. DLA can implement a neural network that regresses the confidence values. The neural network may receive as its inputs at least some subset of parameters, such as bounding box dimensions, ground plane estimates obtained (e.g., from another subsystem), inertial measurement unit (IMU) sensor 1366 outputs that correlate with the vehicle 1300 orientation, distance, and 3D position estimates of the object obtained from the neural network and / or other sensors (e.g., LIDAR sensor 1364 or RADAR sensor 1360), and others.

[0131] The SoC 1304 may include a data store 1316 (e.g., a memory). The data store 1316 may be an on-chip memory of the SoC 1304 and may store the neural network to be executed by the GPU and / or DLA. In some instances, the data store 1316 may have a capacity large enough to store multiple instances of the neural network for redundancy and safety. The data store 1316 may comprise an L2 or L3 cache 1312. References to the data store 1316 may include references to memory associated with the PVA, DLA, and / or other accelerators 1314, as described herein.

[0132] The SoC 1304 may include one or more processors 1310 (e.g., embedded processors). The processors 1310 may include a boot and power management processor, which may be a dedicated processor and subsystem for handling boot power and management capabilities and associated security enforcement. The boot and power management processor may be part of the SoC 1304 boot sequence and may provide run-time power management services. The boot power and management processor may provide clock and voltage programming, assist with system low power state transitions, management of the SoC 1304 thermal and temperature sensors, and / or management of the SoC 1304 power states. Each temperature sensor may be implemented as a ring oscillator whose output frequency is proportional to temperature, and the SoC 1304 may use the ring oscillator to detect the temperature of the CPU 1306, the GPU 1308, and / or the accelerator 1314. If the temperature is determined to exceed a threshold, the boot and power management processor may enter a temperature fault routine, place the SoC 1304 in a lower power state and / or place the vehicle 1300 in a Chauffeur safe shutdown mode (e.g., bring the vehicle 1300 to a safe shutdown).

[0133] The processor 1310 may further include a set of embedded processors that may perform the functions of an audio processing engine. The audio processing engine may be an audio subsystem that allows full hardware support of multi-channel audio through multiple interfaces and a wide and flexible range of audio I / O interfaces. In some instances, the audio processing engine is a dedicated processor core that includes a digital signal processor with dedicated RAM.

[0134] The processor 1310 may further include an always-on processor engine that can provide the necessary hardware features to support low power sensor management and wake use cases. The always-on processor engine may include a processor core, tightly coupled RAM, support peripherals (e.g., timers and interrupt controllers), various I / O controller peripherals, and routing logic.

[0135] The processor 1310 may further include a safety cluster engine that includes a processor subsystem dedicated to handling safety management of automotive applications. The safety cluster engine may include two or more processor cores, tightly coupled RAM, supporting peripherals (e.g., timers, interrupt controllers, etc.), and / or routing logic. In a safety mode, the two or more cores may operate in lockstep mode and function as a single core with comparison logic to detect any differences between their operations.

[0136] The processor 1310 may further include a real-time camera engine, which may include a dedicated processor subsystem for handling real-time camera management.

[0137] The processor 1310 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.

[0138] The processor 1310 may include a video image compositor, which may be a processing block (e.g., implemented in a microprocessor) that implements video post-processing functions required by the video playback application to produce a final image for the player window. The video image compositor may perform lens distortion correction on the wide-view camera 1370, on the surround camera 1374, and / or on the in-cabin surveillance camera sensor. The in-cabin surveillance camera sensor is preferably monitored by a neural network running on another instance of the advanced SoC, configured to identify in-cabin events and respond appropriately. The in-cabin system may perform lip reading to activate cellular service and make phone calls, dictate emails, change the destination of the vehicle, activate or change the vehicle's infotainment system and settings, or provide voice-activated web surfing. Certain features are only available to the driver when operating in autonomous mode and are disabled otherwise.

[0139] The video image synthesizer may include enhanced temporal noise reduction for both spatial and temporal noise reduction. For example, when motion occurs in the video, the noise reduction reduces the weight of information provided by adjacent frames and appropriately weights the spatial information. When an image or part of an image does not contain motion, the temporal noise reduction performed by the video image synthesizer may use information from previous images to reduce noise in the current image.

[0140] The video image composer may also be configured to perform stereo rectification on the input stereo lens frames. The video image composer may further be used for user interface compositing when the operating system desktop is in use and the GPU 1308 is not required to continuously render new surfaces. Even when the GPU 1308 is powered on and actively performing 3D rendering, the video image composer may be used to offload the GPU 1308 for improved performance and responsiveness.

[0141] The SoC1304 may further include a mobile industry processor interface (MIPI) camera serial interface, a high-speed interface for receiving video and input from a camera, and / or a video input block that may be used for camera and related pixel input functions. The SoC1304 may further include an input / output controller that may be controlled by software and that may be used to receive I / O signals that are not committed to a specific role.

[0142] The SoC 1304 may further include a wide range of peripheral interfaces to enable communication with peripherals, audio codecs, power management, and / or other devices. The SoC 1304 may be used to process data from cameras (e.g., connected via gigabit multimedia serial links and Ethernet), sensors (e.g., LIDAR sensor 1364, RADAR sensor 1360, etc., which may be connected via Ethernet), data from the bus 1302 (e.g., vehicle 1300 speed, steering wheel position, etc.), and GNSS sensor 1358 (e.g., connected via Ethernet or CAN bus). The SoC 1304 may further include a dedicated high-performance mass storage controller, which may include its own DMA engine and may be used to offload the CPU 1306 from routine data management tasks.

[0143] The SoC1304 may be an end-to-end platform with a flexible architecture that spans levels 3-5 of automation, thereby providing a comprehensive functional safety architecture that leverages and efficiently uses computer vision and ADAS techniques for diversity and redundancy, and provides a platform for a flexible, reliable driving software stack along with deep learning tools. The SoC1304 may be faster, more reliable, and more energy- and space-efficient than traditional systems. For example, when the accelerator 1314 is combined with the CPU 1306, the GPU 1308, and the data store 1316, the SoC1304 may provide a fast and efficient platform for levels 3-5 of autonomous vehicles.

[0144] Thus, the present technology provides capabilities and functionality that cannot be achieved by conventional systems. For example, computer vision algorithms may be executed on a CPU, which may be configured using a high-level programming language, such as the C programming language, to execute a wide variety of processing algorithms across a wide variety 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 in-vehicle ADAS applications and for practical Level 3-5 autonomous vehicles.

[0145] In contrast to conventional systems, by providing a CPU complex, a GPU complex, and a hardware acceleration cluster, the techniques described herein allow multiple neural networks to run simultaneously and / or sequentially and the results to be combined to enable Level 3-5 autonomous driving capabilities. For example, a CNN running on the DLA or dGPU (e.g., GPU1320) can include text and word recognition that allows the supercomputer to read and understand traffic signs, including signs for which the neural network was not specifically trained. The DLA can further include neural networks that can identify, interpret, and provide a semantic understanding of the signs and pass the semantic understanding to a route planning module running on the CPU complex.

[0146] As another example, multiple neural networks may be run simultaneously, as required for level 3, 4, or 5 operation. For example, a warning sign consisting of "Caution: Flashing lights indicate icy conditions" along with a lightning flash may be interpreted by several neural networks, independently or collectively. The sign itself may be identified as a traffic sign by a first deployed neural network (e.g., a trained neural network), and the text "Flashing lights indicate icy conditions" may be interpreted by a second deployed neural network that informs the vehicle's route planning software (preferably running on a CPU complex) that icy conditions exist when a flashing light is detected. The flashing light may be identified by running a third deployed neural network through multiple frames, informing the vehicle's route planning software of the presence (or absence) of the flashing light. All three neural networks may be run simultaneously, such as within the DLA and / or on the GPU 1308.

[0147] In some instances, a CNN for facial recognition and vehicle owner identification can use data from the camera sensor to identify the presence of a legitimate driver and / or owner of the vehicle 1300. The always-on sensor processing engine can be used to unlock the vehicle and turn on the lights when the owner approaches the driver's door, and in security mode, to stop operation of the vehicle when the owner leaves the vehicle. In this manner, the SoC 1304 provides security against theft and / or vehicle takeover.

[0148] In another example, a CNN for emergency vehicle detection and identification can detect and identify emergency vehicle sirens using data from the microphone 1396. In contrast to conventional systems that use general classifiers to detect sirens and manually extract features, the SoC 1304 uses CNNs for classification of environmental and urban sounds, as well as classification of visual data. In a preferred embodiment, the CNN running on the DLA is trained to identify the relative terminal velocity of emergency vehicles (e.g., by using the Doppler effect). The CNN can also be trained to identify emergency vehicles specific to the local area in which the mobile vehicle is operating, as identified by the GNSS sensor 1358. Thus, for example, when operating in Europe, the CNN will attempt to detect European sirens, and when in the United States, the CNN will attempt to identify only North American sirens. After an emergency vehicle is detected, the control program may be used to execute emergency vehicle safety routines, such as slowing down the mobile vehicle, stopping the mobile vehicle at the side of the road, parking the mobile vehicle, and / or idling the mobile vehicle, with the assistance of ultrasonic sensor 1362, until the emergency vehicle has passed.

[0149] The vehicle may include a CPU 1318 (e.g., a separate CPU, or a dCPU) that may be coupled to the SoC 1304 via a high-speed interconnect (e.g., PCIe). The CPU 1318 may include, for example, an X86 processor. The CPU 1318 may be used to perform any of a variety of functions including, for example, reconciling potentially inconsistent results between the ADAS sensors and the SoC 1304 and / or monitoring the status and health of the controller 1336 and / or the infotainment SoC 1330.

[0150] The vehicle 1300 may include a GPU 1320 (e.g., a discrete GPU, or a dGPU) that may be coupled to the SoC 1304 via a high-speed interconnect (e.g., NVIDIA's NVLINK). The GPU 1320 may provide additional artificial intelligence functionality, such as by running redundant and / or different neural networks, and may be used to train and / or update neural networks based on input (e.g., sensor data) from sensors of the vehicle 1300.

[0151] The mobile vehicle 1300 may further include a network interface 1324, which may include one or more wireless antennas 1326 (e.g., one or more wireless antennas for different communication protocols, such as a cellular antenna, a Bluetooth antenna, etc.). The network interface 1324 may be used to enable wireless connections with the cloud over the Internet (e.g., with the server 1378 and / or other network devices), with other mobile vehicles, and / or with computing devices (e.g., passenger client devices). To communicate with other mobile vehicles, a direct link may be established between the two mobile vehicles and / or an indirect link may be established (e.g., through a network and via the Internet). A direct link may be provided using a mobile vehicle-to-mobile communication link. The mobile vehicle-to-mobile communication link may provide the mobile vehicle 1300 information about mobile vehicles in its vicinity (e.g., mobile vehicles in front of, beside, and / or behind the mobile vehicle 1300). This feature may be part of a collaborative adaptive cruise control feature of the vehicle 1300 .

[0152] The network interface 1324 may include a SoC that provides modulation and demodulation functions and enables the controller 1336 to communicate over a wireless network. The network interface 1324 may include a radio frequency front end for upconversion from baseband to radio frequency and downconversion from radio frequency to baseband. The frequency conversion may be performed through well known processes and / or may be performed using a superheterodyne process. In some instances, the radio frequency front end functions may be provided by a separate chip. The network interface may include wireless functions for communicating via LTE, WCDMA, UMTS, GSM, CDMA2000, Bluetooth, Bluetooth LE, Wi-Fi, Z-Wave, ZigBee, LoRaWAN, and / or other wireless protocols.

[0153] Mobile vehicle 1300 may further include a data store 1328, which may include off-chip (e.g., off-SoC 1304) storage. Data store 1328 may include one or more memory elements, including RAM, SRAM, DRAM, VRAM, flash, hard disk, and / or other components and / or devices capable of storing at least one bit of data.

[0154] The vehicle 1300 may further include GNSS sensors 1358 (e.g., GPS and / or assisted GPS sensors) to assist in mapping, perception, occupancy grid generation, and / or route planning functions. Any number of GNSS sensors 1358 may be used, including, but not limited to, a GPS using a USB connector with an Ethernet to serial (RS-232) bridge.

[0155] The mobile vehicle 1300 may further include a RADAR sensor 1360. The RADAR sensor 1360 may be used by the mobile vehicle 1300 for long-range mobile vehicle detection, even in darkness and / or severe weather conditions. The RADAR functional safety level may be ASIL B. In some instances, the RADAR sensor 1360 may use the CAN and / or bus 1302 for control and to access object tracking data (e.g., to transmit data generated by the RADAR sensor 1360), with access to Ethernet to access raw data. A wide variety of RADAR sensor types may be used. For example, and without limitation, the RADAR sensor 1360 may be suitable for front, rear, and side RADAR use. In some instances, a pulsed Doppler RADAR sensor is used.

[0156] The RADAR sensor 1360 may include different configurations, such as long range with a narrow field of view, short range with a wide field of view, and short range side coverage. In some instances, the long range RADAR may be used for adaptive cruise control functions. Long range RADAR systems may provide a wide field of view achieved by two or more independent scans, such as within a 250m range. The RADAR sensor 1360 may help distinguish between static and moving objects and may be used by ADAS systems for emergency brake assist and forward collision warning. Long range RADAR sensors may include monostatic multimodal RADAR with multiple (e.g., six or more) fixed RADAR antennas and high speed CAN and FlexRay interfaces. In one instance with six antennas, the center four antennas may create a focused beam pattern designed to record the surroundings of the moving vehicle 1300 at high speeds with minimal interference from traffic in adjacent lanes. The other two antennas may widen the field of view, allowing for rapid detection of moving vehicles entering or leaving the lane of the moving vehicle 1300.

[0157] As an example, a medium-range RADAR system may include a range of up to 1360m (front) or 80m (rear) and a field of view of up to 42 degrees (front) or 1350 degrees (rear). A short-range RADAR system may include, but is not limited to, a RADAR sensor designed to be mounted on either end of a rear bumper. When mounted on either end of a rear bumper, such a RADAR sensor system can create two beams that constantly monitor the blind spots behind and adjacent to a moving vehicle.

[0158] Short-range RADAR systems may be used in ADAS systems for blind spot detection and / or lane change assist.

[0159] The mobile vehicle 1300 may further include ultrasonic sensors 1362. The ultrasonic sensors 1362, which may be located at the front, rear, and / or sides of the mobile vehicle 1300, may be used for parking assist and / or for creating and updating an occupancy grid. A wide variety of ultrasonic sensors 1362 may be used, with different ultrasonic sensors 1362 being used for different ranges of detection (e.g., 2.5m, 4m). The ultrasonic sensors 1362 may operate at a functional safety level of ASIL B.

[0160] The mobile vehicle 1300 may include a LIDAR sensor 1364. The LIDAR sensor 1364 may be used for object and pedestrian detection, emergency braking, collision avoidance, and / or other functions. The LIDAR sensor 1364 may be functional safety level ASIL B. In some instances, the mobile vehicle 1300 may include multiple (e.g., 2, 4, 6, etc.) LIDAR sensors 1364 that may use Ethernet (e.g., to provide data to a Gigabit Ethernet switch).

[0161] In some instances, the LIDAR sensor 1364 may have the ability to provide a list of objects and their distances in a 360-degree field of view. A commercially available LIDAR sensor 1364 may have an advertised range of about 100 m, for example, with an accuracy of 2 cm to 3 cm, and with support for a 100 Mbps Ethernet connection. In some instances, one or more non-protruding LIDAR sensors 1364 may be used. In such instances, the LIDAR sensor 1364 may be implemented as a small device that may be integrated into the front, rear, sides, and / or corners of the mobile vehicle 1300. In such instances, the LIDAR sensor 1364 may have a range of 200 m even for low-reflecting objects, and may provide up to a 120-degree horizontal and 35-degree vertical field of view. A front-mounted LIDAR sensor 1364 may be configured for a horizontal field of view between 45 and 135 degrees.

[0162] In some instances, LIDAR technology such as 3D flash LIDAR may also be used. 3D flash LIDAR uses a laser flash as a transmitter to illuminate the surroundings of the vehicle up to about 200 m. The flash LIDAR unit includes a receptor that records the laser pulse transit time and the reflected light on each pixel, which in turn corresponds to the range from the vehicle to the object. Flash LIDAR may allow a highly accurate and distortion-free image of the surroundings to be generated with every laser flash. In some instances, four flash LIDAR sensors may be deployed, one on each side of the vehicle 1300. Available 3D flash LIDAR systems include solid-state 3D steering array LIDAR cameras (e.g., non-scanning LIDAR devices) that have no moving parts other than the blower. The flash LIDAR device may use a 5 nanosecond Class I (eye-safe) laser pulse per frame and may capture the reflected laser light in the form of a 3D range point cloud and coregistered intensity data. By using flash LIDAR, and because flash LIDAR is a solid-state device with no moving parts, the LIDAR sensor 1364 may be less susceptible to motion blur, vibration, and / or shock.

[0163] The vehicle may further include an IMU sensor 1366. In some instances, the IMU sensor 1366 may be positioned at the center of the rear axle of the vehicle 1300. The IMU sensor 1366 may include, for example, but not limited to, an accelerometer, a magnetometer, a gyroscope, a magnetic compass, and / or other sensor types. In some instances, such as in a 6-axis application, the IMU sensor 1366 may include an accelerometer and a gyroscope, while in a 9-axis application, the IMU sensor 1366 may include an accelerometer, a gyroscope, and a magnetometer.

[0164] In some embodiments, the IMU sensor 1366 may be implemented as a miniature, high-performance GPS-Aided Inertial Navigation System (GPS / INS) that combines micro-electro-mechanical system (MEMS) inertial sensors, highly sensitive GPS receivers, and advanced Kalman filtering algorithms to provide estimates of position, velocity, and attitude. As such, in some instances, the IMU sensor 1366 may enable the vehicle 1300 to estimate heading without requiring input from a magnetic sensor by directly observing and correlating changes in velocity from the GPS to the IMU sensor 1366. In some instances, the IMU sensor 1366 and the GNSS sensor 1358 may be combined in a single integrated unit.

[0165] The vehicle may include microphones 1396 placed in and / or around the vehicle 1300. The microphones 1396 may be used for emergency vehicle detection and identification, among other things.

[0166] The vehicle may further include any number of camera types, including stereo cameras 1368, wide view cameras 1370, infrared cameras 1372, surround cameras 1374, long and / or mid-range cameras 1398, and / or other camera types. The cameras may be used to capture image data around the entire exterior of the vehicle 1300. The type of cameras used depends on the embodiment and requirements of the vehicle 1300, and any combination of camera types may be used to achieve the required coverage around the vehicle 1300. Additionally, 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. The cameras may support, by way of example only, Gigabit Multimedia Serial Link (GMSL) and / or Gigabit Ethernet, but are not limited thereto. Each camera is described in further detail herein in conjunction with Figures 13A and 13B.

[0167] The mobile vehicle 1300 may further include a vibration sensor 1342. The vibration sensor 1342 may measure vibrations of components of the mobile vehicle, such as an axle. For example, a change in vibration may indicate a change in the surface of the road. In another example, when two or more vibration sensors 1342 are used, the difference in vibration may be used to determine friction or slippage of the road surface (e.g., when the difference in vibration is between a powered axle and a free-running axle).

[0168] The mobile vehicle 1300 may include an ADAS system 1338. In some instances, the ADAS system 1338 may include a SoC. The ADAS system 1338 may include autonomous / adaptive / automatic cruise control (ACC), cooperative adaptive cruise control (CACC), forward crash warning (FCW), automatic emergency braking (AEB), lane departure warning (LDW), lane keep assist (LKA), blind spot warning (BSW), rear cross-traffic warning (RCTW), collision warning system (CWS), lane centering (LC), and / or other features and functions.

[0169] The ACC system may use a RADAR sensor 1360, a LIDAR sensor 1364, and / or a camera. The ACC system may include longitudinal ACC and / or lateral ACC. The longitudinal ACC monitors and controls the distance to the vehicle directly ahead of the vehicle 1300 and automatically adjusts the vehicle speed to maintain a safe distance from the vehicle ahead. The lateral ACC performs distance keeping and advises the vehicle 1300 to change lanes when necessary. The lateral ACC is related to other ADAS applications such as LC and CWS.

[0170] CACC uses information from other moving vehicles, which may be received from other moving vehicles via a wireless link via the network interface 1324 and / or the wireless antenna 1326, or indirectly via a network connection (e.g., via the Internet). A direct link may be provided by a vehicle-to-vehicle (V2V) communication link, while an indirect link may be an infrastructure-to-vehicle (I2V) communication link. In general, a V2V communication concept provides information about the immediately preceding moving vehicle (e.g., the moving vehicle directly ahead of the moving vehicle 1300, in the same lane as the moving vehicle 1300), while an I2V communication concept provides information about the traffic further ahead. A CACC system may include either or both I2V and V2V information sources. Given the information of moving vehicles ahead of the moving vehicle 1300, CACC may be more reliable and may potentially make traffic flow smoother and reduce congestion on the roads.

[0171] The FCW system is designed to warn the driver of hazards so that the driver can take corrective action. The FCW system uses a forward-facing camera and / or RADAR sensor 1360 coupled to a dedicated processor, DSP, FPGA, and / or ASIC, electrically coupled to driver feedback such as a display, speaker, and / or vibrating components. The FCW system can provide warnings in the form of audio, visual alerts, vibrations, and / or quick brake pulses.

[0172] An AEB system can detect an imminent forward collision with another moving vehicle or other object and automatically apply the brakes if the driver does not take corrective action within specified time or distance parameters. The AEB system can use a forward-facing camera and / or RADAR sensor 1360 coupled to a dedicated processor, DSP, FPGA, and / or ASIC. When the AEB system detects a hazard, the AEB system typically first alerts the driver to take corrective action to avoid the collision, and if the driver does not take corrective action, the AEB system can automatically apply the brakes as part of an effort to prevent or at least mitigate the effects of the predicted collision. The AEB system can include techniques such as dynamic brake support and / or collision imminent braking.

[0173] The LDW system provides visual, audible, and / or tactile warnings, such as vibrations of the steering wheel or seat, to alert the driver when the mobile vehicle 1300 crosses a lane marking. The LDW system is not activated when the driver indicates an intentional lane departure by activating a turn signal. The LDW system may use a forward-facing camera coupled to a dedicated processor, DSP, FPGA, and / or ASIC electrically coupled to driver feedback, such as a display, speaker, and / or vibrating components.

[0174] The LKA system is a modification of the LDW system, which provides steering input or braking to correct the vehicle 1300 if it begins to wander out of its lane.

[0175] The BSW system detects and warns the driver of a moving vehicle in the vehicle's blind spot. The BSW system can provide visual, audible, and / or tactile warnings to indicate that merging or changing lanes is unsafe. The system can provide additional warnings when the driver uses turn signals. The BSW system can use rear-facing cameras and / or RADAR sensors 1360 coupled to dedicated processors, DSPs, FPGAs, and / or ASICs electrically coupled to driver feedback, e.g., displays, speakers, and / or vibration components.

[0176] The RCTW system may provide visual, audible, and / or tactile notifications when an object is detected outside the range of the rear camera when the vehicle 1300 is backing up. Some RCTW systems include AEB to ensure vehicle brakes are applied to avoid a collision. The RCTW system may use one or more rear-facing RADAR sensors 1360 coupled to a dedicated processor, DSP, FPGA, and / or ASIC electrically coupled to driver feedback, e.g., a display, speaker, and / or vibration components.

[0177] Because conventional ADAS systems alert the driver and allow the driver to determine whether a safety condition truly exists and act accordingly, conventional ADAS systems can be prone to producing false positive results that are usually not catastrophic but can be annoying and distracting to the driver. However, in an autonomous vehicle 1300, the vehicle 1300 itself must decide whether to listen to the results from the primary computer or the secondary computer (e.g., the first controller 1336 or the second controller 1336) when the results are conflicting. For example, in some embodiments, the ADAS system 1338 may be a backup and / or secondary computer to provide perception information to a backup computer rationality module. The backup computer rationality monitor can run redundant and diverse software on hardware components to detect impairments in perception and dynamic driving tasks. Output from the ADAS system 1338 may be provided to a supervisory MCU. When the outputs from the primary and secondary computers conflict, the supervisory MCU must decide how to reconcile the conflict to ensure safe operation.

[0178] In some instances, the primary computer may be configured to provide a reliability score to the overseer MCU indicating the reliability of the primary computer in a selected outcome. If the reliability score exceeds a threshold, the overseer MCU may follow the instructions of the primary computer regardless of whether the secondary computers provide conflicting or inconsistent results. If the reliability score does not meet the threshold, and if the primary and secondary computers show different (e.g., conflicting) results, the overseer MCU may arbitrate between the computers to determine the appropriate outcome.

[0179] The supervisory MCU may be configured to execute a neural network that is trained and configured to determine conditions under which the secondary computer provides a false alarm based on the outputs from the primary and secondary computers. Thus, the neural network in the supervisory MCU may learn when the output of the secondary computer may be trusted and when it may not be trusted. For example, when the secondary computer is a RADAR-based FCW system, the neural network in the supervisory MCU may learn when the FCW identifies a metal object that is not actually a hazard, such as a sewer grate or manhole cover, that triggers an alarm. Similarly, when the secondary computer is a camera-based LDW system, the neural network in the supervisory MCU may learn to ignore the LDW when a bicyclist or pedestrian is present and lane departure is, in fact, the safest maneuver. In embodiments that include a neural network executing on the supervisory MCU, the supervisory MCU may include at least one of a DLA or a GPU suitable for executing the neural network with associated memory. In a preferred embodiment, the director MCU may comprise and / or be included as a component of the SoC1304.

[0180] In other instances, the ADAS system 1338 may include a secondary computer that performs ADAS functions using classical rules of computer vision. As such, the secondary computer may use classical computer vision rules (if-then) and the presence of a neural network in the supervisory MCU may improve reliability, safety and performance. For example, the diverse implementations and intentional non-identity may make the overall system more fault-tolerant, especially to failures caused by software (or software-hardware interface) functions. For example, if there is a software bug or error in the software running on the primary computer and non-identical software code running on the secondary computer provides the same overall result, the supervisory MCU may have greater confidence that the overall result is correct and that a bug in the software or hardware used by the primary computer has not caused a critical error.

[0181] In some instances, the output of the ADAS system 1338 may be fed to a perception block of the primary computer and / or a dynamic driving task block of the primary computer. For example, if the ADAS system 1338 indicates a forward collision warning due to an object directly ahead, the perception block may use this information when identifying the object. In other instances, the secondary computer may have its own neural network that is trained as described herein, thus reducing the risk of false positives.

[0182] The mobile vehicle 1300 may further include an infotainment SoC 1330 (e.g., an in-vehicle infotainment system (IVI)). Although shown and described as an SoC, the infotainment system need not be an SoC and may include two or more separate components. The infotainment SoC 1330 may include a combination of hardware and software that may be used to provide audio (e.g., music, personal digital assistant, navigation instructions, news, radio, etc.), video (e.g., TV, movies, streaming, etc.), telephony (e.g., hands-free calling), network connectivity (e.g., LTE, Wi-Fi, etc.), and / or information services (e.g., navigation system, rear parking assist, wireless data system, mobile vehicle related information such as fuel level, total distance traveled, brake fuel level, oil level, door opening / closing, air filter information, etc.) to the mobile vehicle 1300. For example, the infotainment SoC 1330 may include a radio, a disc player, a navigation system, a video player, USB and Bluetooth connectivity, a car computer, in-car entertainment, Wi-Fi, steering wheel audio controls, hands-free voice control, a heads-up display (HUD), an HMI display 1334, telematics devices, a control panel (e.g., for controlling and / or interacting with various components, features, and / or systems), and / or other components. The infotainment SoC 1330 may further be used to provide information (e.g., visual and / or audible) to a user of the vehicle, such as information from an ADAS system 1338, autonomous driving information such as planned vehicle maneuvers, trajectory, surrounding environment information (e.g., intersection information, vehicle information, road information, etc.), and / or other information.

[0183] The infotainment SoC 1330 may include GPU functionality. The infotainment SoC 1330 may communicate with other devices, systems, and / or components of the mobile vehicle 1300 via the bus 1302 (e.g., CAN bus, Ethernet, etc.). In some instances, the infotainment SoC 1330 may be coupled to a supervisory MCU such that the infotainment system's GPU can perform some self-drive functions in the event that the primary controller 1336 (e.g., the primary and / or backup computer of the mobile vehicle 1300) fails. In such instances, the infotainment SoC 1330 can place the mobile vehicle 1300 in a Chauffeur safe stop mode, as described herein.

[0184] The mobile vehicle 1300 may further include an instrument cluster 1332 (e.g., a digital dash, an electronic instrument cluster, a digital instrument panel, etc.). The instrument cluster 1332 may include a controller and / or a supercomputer (e.g., a separate controller or a supercomputer). The instrument cluster 1332 may include a set of instruments such as a speedometer, a fuel level, an oil pressure, a tachometer, an odometer, a turn signal, a gear shift position indicator, a seat belt warning light, a parking brake warning light, an engine malfunction light, an airbag (SRS) system information, lighting controls, safety system controls, navigation information, etc. In some instances, information may be displayed and / or shared between the infotainment SoC 1330 and the instrument cluster 1332. In other words, the instrument cluster 1332 may be included as part of the infotainment SoC 1330, or vice versa.

[0185] 13D is a system diagram of communication between a cloud-based server and an example autonomous vehicle 1300 of FIG. 13A in accordance with some embodiments of the present disclosure. System 1376 may include a server 1378, a network 1390, and a mobile vehicle including mobile vehicle 1300. Server 1378 may include multiple GPUs 1384(A)-1384(H) (collectively referred to herein as GPUs 1384), PCIe switches 1382(A)-1382(H) (collectively referred to herein as PCIe switches 1382), and / or CPUs 1380(A)-1380(B) (collectively referred to herein as CPUs 1380). GPUs 1384, CPUs 1380, and PCIe switches may be interconnected with a high-speed interconnect, such as, but not limited to, an NVLink interface 1388 developed by NVIDIA and / or a PCIe connection 1386. In some instances, the GPUs 1384 are connected via an NVLink and / or NVSwitch SoC, and the GPUs 1384 and PCIe switch 1382 are connected via a PCIe interconnect. Although eight GPUs 1384, two CPUs 1380, and two PCIe switches are illustrated, this is not intended to be limiting. Depending on the embodiment, each server 1378 may include any number of GPUs 1384, CPUs 1380, and / or PCIe switches. For example, the servers 1378 may each include 8, 16, 32, and / or more GPUs 1384.

[0186] The server 1378 can receive image data from the mobile vehicle over the network 1390, the image data representing images showing unexpected or changed road conditions, such as recently started road construction. The server 1378 can transmit the neural network 1392, the updated neural network 1392, and / or the map information 1394, including information about traffic and road conditions, to the mobile vehicle over the network 1390. The updates to the map information 1394 can include updates to the HD map 1322, such as information about construction sites, potholes, detours, flooding, and / or other obstacles. In some instances, the neural network 1392, the updated neural network 1392, and / or the map information 1394 may have arisen from new training and / or experience represented in data received from any number of mobile vehicles in the environment and / or based on training performed at a data center (e.g., using the server 1378 and / or other servers).

[0187] The server 1378 may be used to train a machine learning model (e.g., a neural network) based on the training data. The training data may be generated by a moving vehicle and / or generated in a simulation (e.g., using a game engine). In some instances, the training data is tagged (e.g., if the neural network benefits from supervised learning) and / or undergoes other pre-processing, while in other instances, the training data is not tagged and / or pre-processed (e.g., if the neural network does not require supervised learning). The training may be performed according to any one or more classes of machine learning techniques, including, but not limited to, the following classes: supervised training, semi-supervised training, unsupervised training, self-learning, reinforcement learning, federated learning, transfer learning, feature learning (including principal component and cluster analysis), multi-linear subspace learning, manifold learning, representation learning (including preliminary dictionary learning), rule-based machine learning, anomaly detection, and variations or combinations thereof. After the machine learning model has been traced, the machine learning model may be used by the vehicle (e.g., transmitted to the vehicle via network 1390) and / or the machine learning model may be used by server 1378 to remotely monitor the vehicle.

[0188] In some instances, the server 1378 can receive data from the moving vehicles and apply the data to state-of-the-art real-time neural networks for real-time intelligent inference. The server 1378 can include deep learning supercomputers and / or dedicated AI computers powered by the GPU 1384, such as the DGX and DGX Station machines developed by NVIDIA. However, in some instances, the server 1378 can include a deep learning infrastructure that uses only CPU-powered data centers.

[0189] The deep learning infrastructure of the server 1378 may be capable of rapid real-time inference and may use that capability to evaluate and validate the health of the processor, software, and / or associated hardware within the moving vehicle 1300. For example, the deep learning infrastructure may receive periodic updates from the moving vehicle 1300, such as sequences of images and / or objects that the moving vehicle 1300 has located within the images of that sequence (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 the moving vehicle 1300, and if the results do not match and the infrastructure concludes that the AI ​​within the moving vehicle 1300 is not functioning properly, the server 1378 may send a signal to the moving vehicle 1300 to infer control, notify passengers, and command the failsafe computer of the moving vehicle 1300 to complete a safe parking maneuver.

[0190] For inference, the server 1378 may include a GPU 1384 and one or more programmable inference accelerators (e.g., NVIDIA's TensorRT). The combination of a GPU-powered server and inference acceleration can enable real-time responsiveness. In other instances, such as when less performance is required, servers powered by CPUs, FPGAs, and other processors may be used for inference.

[0191] Exemplary Computing Device 14 is a block diagram of an example computing device 1400 suitable for use in implementing some embodiments of the present disclosure. The computing device 1400 may include an interconnection system 1402 that directly or indirectly couples the following devices: memory 1404, one or more central processing units (CPUs) 1406, one or more graphic processing units (GPUs) 1408, a communication interface 1410, input / output (I / O) ports 1412, input / output components 1414, a power supply 1416, one or more presentation components 1418 (e.g., displays), and one or more logic units 1420.

[0192] Although the various blocks in FIG. 14 are shown connected via interconnect system 1402 with lines, this is not intended to be limiting and is merely for clarity. For example, in some embodiments, a presentation component 1418, such as a display device, may be considered an I / O component 1414 (e.g., where the display is a touch screen). As another example, CPU 1406 and / or GPU 1408 may include memory (e.g., memory 1404 may represent a storage device in addition to the memory of GPU 1408, CPU 1406, and / or other components). In other words, the computing devices of FIG. 14 are merely exemplary. Categories such as “workstations,” “servers,” “laptops,” “desktops,” “tablets,” “client devices,” “mobile devices,” “handheld devices,” “gaming consoles,” “electronic control units (ECUs),” “virtual reality systems,” “augmented reality systems,” and / or other device or system types are all intended to be within the scope of the computing devices of FIG. 14 and therefore are not to be differentiated.

[0193] The interconnect system 1402 may represent one or more links or buses, such as an address bus, a data bus, a control bus, or a combination thereof. The interconnect system 1402 may include one or more bus or link types, such as an industry standard architecture (ISA) bus, an extended industry standard architecture (EISA) bus, a video electronics standards association (VESA) bus, a peripheral component interconnect (PCI) bus, a peripheral component interconnect express (PCIe) bus, and / or another type of bus or link. In some embodiments, there are direct connections between components. As an example, the CPU 1406 may be directly connected to the memory 1404. Additionally, the CPU 1406 may be directly connected to the GPU 1408. When there is a direct or point-to-point connection between components, the interconnect system 1402 may include a PCIe link to implement the connection. In these instances, a PCI bus need not be included in the computing device 1400 .

[0194] Memory 1404 may include any of a variety of computer readable media. Computer readable media may be any available media that can be accessed by computing device 1400. Computer readable media may include both volatile and nonvolatile media, and removable and non-removable media. By way of example, and not limitation, computer readable media may comprise computer storage media and communication media.

[0195] Computer storage media may include both volatile and nonvolatile media and / or removable and non-removable media implemented in any method or technology for storage of information, such as computer readable instructions, data structures, program modules, and / or other data types. For example, memory 1404 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 memory technology, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium that may be used to store the desired information and that may be accessed by computing device 1400. As used herein, computer storage media does not include the signals themselves.

[0196] Computer storage media may embody computer-readable instructions, data structures, program modules, and / or other data types in a modulated data signal such as a carrier wave or other transport mechanism and include any information delivery media. The term "modulated data signal" may refer to a signal that has one or more of its characteristics set or changed in such a manner as to encode information in the signal. By way of example, and not limitation, computer storage media may include wired media, such as a wired network or direct-wired connection, and wireless media, such as acoustic, RF, infrared and other wireless media. Combinations of any of the above should also be included within the scope of computer-readable media.

[0197] The CPU 1406 may be configured to execute at least some of the computer readable instructions to control one or more components of the computing device 1400 to perform one or more of the methods and / or processes described herein. The CPU 1406 may include one or more (e.g., 1, 2, 4, 8, 28, 72, etc.) cores each capable of processing multiple software threads simultaneously. The CPU 1406 may include any type of processor, and may include different types of processors depending on the type of computing device 1400 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 1400, the processor may be an Advanced RISC Machines (ARM) processor implemented using Reduced Instruction Set Computing (RISC) or an x86 processor implemented using Complex Instruction Set Computing (CISC). The computing device 1400 may include one or more CPUs 1406 within one or more microprocessors or auxiliary coprocessors, such as computational coprocessors.

[0198] In addition to or in lieu of the CPU 1406, the GPU 1408 may be configured to execute at least some of the computer readable instructions to control one or more components of the computing device 1400 to perform one or more of the methods and / or processes described herein. One or more of the GPUs 1408 may be integrated GPUs (e.g., with one or more of the CPUs 1406 and / or one or more of the GPUs 1408 may be discrete GPUs. In an embodiment, one or more of the GPUs 1408 may be a coprocessor or coprocessors of the CPUs 1406. The GPU 1408 may be used by the computing device 1400 to render graphics (e.g., 3D graphics) or perform general purpose computations. For example, the GPU 1408 may be configured to perform general purpose computing on a GPU (GPGPU). A graphics processing unit (GPU) 1408 may be used for processing graphics data. The GPU 1408 may include hundreds or thousands of cores capable of processing hundreds or thousands of software threads simultaneously. The GPU 1408 may generate pixel data for an output image in response to rendering commands (e.g., rendering commands from the CPU 1406 received via a host interface). The GPU 1408 may include graphics memory, e.g., display memory, for storing the pixel data or any other suitable data, e.g., GPGPU data. The display memory may be included as part of the memory 1404. GPU 1408 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 may connect the GPUs via a switch (e.g., using NVSwitch). When coupled together, each GPU 1408 may generate pixel data or GPGPU data for a different portion of the output or for a different output (e.g., a first GPU for a first image and a second GPU for a second image). Each GPU may include its own memory or may share memory with the other GPUs.

[0199] In addition to or instead of the CPU 1406 and / or GPU 1408, the logic unit 1420 may be configured to execute at least some of the computer readable instructions to control one or more components of the computing device 1400 to perform one or more of the methods and / or processes described herein. In an embodiment, the CPU 1406, the GPU 1408, and / or the logic unit 1420 may execute any combination of the methods, processes and / or portions thereof, either discretely or jointly. One or more of the logic units 1420 may be part of and / or integrated with one or more of the CPU 1406 and / or GPU 1408, and / or one or more of the logic units 1420 may be a discrete component to or otherwise external to the CPU 1406 and / or GPU 1408. In an embodiment, one or more of logic units 1420 may be a co-processor(s) of one or more of CPUs 1406 and / or one or more of GPUs 1408 .

[0200] Examples of logic unit 1420 include one or more processing cores and / or components thereof, such as a Tensor Core (TC), a Tensor Processing Unit (TPU), a Pixel Visual Core (PVC), a Vision Processing Unit (VPU), a Graphics Processing Cluster (GPC), a Texture Processing Cluster (TPC), a Streaming Multiprocessor (SM), a Tree Traversal Unit (TTU), an Artificial Intelligence Accelerator (AIA), a Deep Learning Accelerator (DLA), an Arithmetic Logic Unit (ALU), an Application Specific Integrated Circuit (ASIC), a Floating Point Unit (FPU), an I / O element, a Peripheral Component Interconnect (PCI) or Peripheral Component Interconnect Express (PCIe) element, and / or the like.

[0201] The communications interface 1410 may include one or more receivers, transmitters, and / or transceivers that enable the computing device 1400 to communicate with other computing devices over electronic communications networks, including wired and / or wireless communications. The communications interface 1410 may include components and functionality to enable communication over any of a number of different networks, such as a wireless network (e.g., Wi-Fi, Z-Wave, Bluetooth, Bluetooth LE, ZigBee, etc.), a wired network (e.g., communicating over Ethernet or InfiniBand), a low power wide area network (e.g., LoRaWAN, SigFox, etc.), and / or the Internet.

[0202] The I / O ports 1412 may enable the computing device 1400 to be logically coupled to other devices, including I / O components 1414, presentation components 1418, and / or other components, some of which may be built-in (e.g., integrated) to the computing device 1400. Exemplary I / O components 1414 include a microphone, a mouse, a keyboard, a joystick, a game pad, a game controller, a satellite dish, a scanner, a printer, a wireless device, and the like. The I / O components 1414 may provide a natural user interface (NUI) that processes air gestures, voice, or other physiological input generated by a user. In some cases, the input may be sent to an appropriate network element for further processing. The NUI may implement any combination of voice recognition, stylus recognition, facial recognition, biometric recognition, gesture recognition on and next to the screen, air gestures, head and eye tracking, and touch recognition in conjunction with the display of the computing device 1400 (as described in more detail below). The computing device 1400 may include a depth camera, such as a stereoscopic camera system, an infrared camera system, an RGB camera system, touch screen technology, and combinations thereof, for gesture detection and recognition. Additionally, the computing device 1400 may include an accelerometer or gyroscope (e.g., as part of an inertia measurement unit (IMU)) to enable detection of motion. In some instances, the output of the accelerometer or gyroscope may be used by the computing device 1400 to render immersive augmented or virtual reality.

[0203] The power supply 1416 may include a hardwired power supply, a battery power supply, or a combination thereof. The power supply 1416 may provide power to the computing device 1400 to enable the components of the computing device 1400 to operate.

[0204] The presentation component 1418 may include a display (e.g., a monitor, a touch screen, a television screen, a heads-up display (HUD), other display types, or a combination thereof), speakers, and / or other presentation components. The presentation component 1418 can receive data from other components (e.g., GPU 1408, CPU 1406, etc.) and output data (e.g., as images, video, sound, etc.).

[0205] The present disclosure may be described in the general context of computer code or machine usable instructions, including computer executable instructions, such as program modules, being executed by a computer or other machine, such as a personal digital assistant or other handheld device. Generally, program modules, including routines, programs, objects, components, data structures, etc., refer to code that performs particular tasks or implements particular abstract data types. The present disclosure may be implemented in a variety of configurations, including handheld devices, consumer electronics, general purpose computers, more specialized computing devices, etc. The present disclosure may also be implemented in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network.

[0206] As used herein, the term "and / or" with respect to two or more elements should be construed to mean only one element or a combination of elements. For example, "element A, element B, and / or element C" may include element A only, element B only, element C only, elements A and element B, elements A and element C, elements B and element C, or elements A, B, and C. In addition, "at least one of element A or element B" may 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. Furthermore, "at least one of element A and element B" may 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.

[0207] The subject matter of the present disclosure has been described with specificity to meet statutory requirements. However, the description itself is not intended to limit the scope of the present disclosure. Instead, the inventors intend that the claimed subject matter may be implemented in other ways, including different steps or combinations of steps similar to those described in this document, in conjunction with other current or future technologies. Furthermore, although the terms "step" and / or "block" may be used herein to connote different elements of the method used, these terms should not be construed as implying any particular order between the various steps disclosed herein unless and when the order of the individual steps is explicitly described.

Claims

1. receiving an image representing a subject and a restraint device corresponding to the subject; performing an initial classification of areas of the image as representing the restraining device; refining the initial classification by applying a set of constraints to the area of ​​the image to obtain a refined classification; generating a model indicative of application of the restraint device based at least in part on the refined classification; Including, A method, wherein the model includes a polynomial curve having coefficients calculated to match a shape of the restraint device.

2. The method of claim 1 , wherein the initial classification is performed by determining a set of pixels that are part of the constraint device using a pixel and a number of neighboring pixels along a specified direction.

3. The method of claim 2 , wherein determining the set of pixels that are part of the constraint device is based at least in part on intensity levels of the pixel and the plurality of neighboring pixels along the specified direction.

4. The method of claim 1 , wherein the restraint device comprises one or more attributes including at least a width size of the restraint device and one or more anchors of the restraint device within a vehicle.

5. The method of claim 1 , wherein the image is captured by the camera with one or more configurations including at least a camera calibration, a camera pose, and a camera lens type.

6. activating a signal to indicate that the restraint device is in an improper location relative to the subject. The method of claim 1 further comprising:

7. determining a location of the restraint device relative to the subject using the model that approximates the shape of the restraint device; The method of claim 1 further comprising:

8. The method of claim 1, further comprising: receiving an image representing a subject and a restraint device corresponding to the subject; performing an initial classification of areas of the image as representing the restraining device; refining the initial classification by applying a set of constraints to the area of ​​the image to obtain a refined classification; generating a model indicative of application of the restraint device based at least in part on the refined classification; Including, A method, wherein the areas of the image correspond to one or more groupings of pixels of the image.

9. 1. A system comprising: one or more processors; a memory for storing computer executable instructions; wherein the computer executable instructions include: performing an initial classification of areas of the image as representing a restraint device applied to the subject; applying a set of constraints to the area of ​​the image to refine the initial classification to obtain a refined classification; and generating a model indicative of application of the restraint device based at least in part on the refined classification. executable by the one or more processors to cause the system to The system, wherein the model includes a polynomial curve having coefficients calculated to match a shape of the restraint device.

10. the system comprises a parallel processing unit (PPU); the image includes one or more areas; 10. The system of claim 9, wherein the one or more areas are classified using one or more threads of the PPU in parallel.

11. The system of claim 9 , wherein the model indicates whether the restraint device is correctly applied to the subject.

12. The system of claim 9 , wherein the set of constraints includes at least a width range.

13. A system comprising: one or more processors; a memory for storing computer executable instructions; wherein the computer executable instructions include: performing an initial classification of areas of the image as representing a restraint device applied to the subject; applying a set of constraints to the area of ​​the image to refine the initial classification to obtain a refined classification; and generating a model indicative of application of the restraint device based at least in part on the refined classification. executable by the one or more processors to cause the system to The area of ​​the image corresponds to one or more groupings of pixels of the image.

14. The system of claim 9 , wherein the images are obtained from one or more networks from one or more systems in a vehicle.

15. The system of claim 14 , wherein the model of the restraint device is provided to the one or more systems of the vehicle via the one or more networks.

16. A vehicle, A propulsion system; an image capture device capable of capturing an image of at least one passenger in the vehicle; A computer system including instructions; and wherein the instructions include at least accessing an image representative of a subject and a restraint device applied to the subject; performing an initial classification of areas of the image as representing the restraining device; refining the initial classification by applying a set of constraints to the area of ​​the image to obtain a refined classification; generating a model indicative of application of the restraint device based at least in part on the refined classification; and controlling a function of at least one subsystem of the vehicle based at least in part on the model. executable by the computer system to perform The vehicle, wherein the model includes a polynomial curve having coefficients calculated to match a shape of the restraint device.

17. 17. The vehicle of claim 16, wherein the restraining device is a seat belt of the vehicle.

18. 17. The vehicle of claim 16, wherein the one subsystem of the vehicle is a warning system that indicates whether the restraint device is worn and correctly applied.

19. 17. The vehicle of claim 16, wherein the one subsystem of the vehicle sends a signal through the propulsion system to control propulsion of the vehicle.

20. 17. The vehicle of claim 16, wherein the one subsystem of the vehicle transmits an indication of whether the restraint device is worn and correctly applied to one or more remote systems.

21. A vehicle, A propulsion system; an image capture device capable of capturing an image of at least one passenger in the vehicle; A computer system including instructions; and wherein the instructions include at least accessing an image representative of a subject and a restraint device applied to the subject; performing an initial classification of areas of the image as representing the restraining device; refining the initial classification by applying a set of constraints to the area of ​​the image to obtain a refined classification; generating a model indicative of application of the restraint device based at least in part on the refined classification; and controlling a function of at least one subsystem of the vehicle based at least in part on the model. executable by the computer system to perform The area of ​​the image corresponds to one or more groupings of pixels of the image.

Images