Abstract: Systems and methods for training artificial intelligence models based on sequences of image data are disclosed. The techniques described herein include generating, using an artificial intelligence model, a respective classification and a respective bounding box for an object depicted in each image of a sequence of images captured during operation of an autonomous vehicle; tracking the object in the sequence of images based on the respective bounding box of each image of the sequence of images and a tracking identifier corresponding to the object; determining a correction to the respective classification of an image of the sequence of images responsive to tracking the object in the sequence of images; and training the artificial intelligence model based on the correction.

Abstract: A system is configured to: (i) perform object detection to identify and detect one or more objects in each image frame captured by each camera of a plurality of cameras; (ii) perform forward tracking and backward tracking for each image frame captured by each camera of the plurality of cameras; (iii) identify and associate objects in each image frame with objects in other frames, for each camera of the plurality of cameras, using a respective bidirectional re-identification (ReID) of each object of the objects; (iv) crop each object of the objects being tracked using the respective bidirectional ReID; (v) generate feature embeddings representing visual characteristics of each cropped object; (vi) generate and update a ranking score for each cropped object based upon the generated feature embeddings for each cropped object; and (vii) based upon the ranking score for each cropped object, associate objects in each image frame.

Abstract: A system is disclosed to: (i) identify a distribution of a first set of datapoints corresponding to a first plurality of objects, of a current training dataset, in a grid including a plurality of bins; (ii) identify a second plurality of objects from an image of a plurality of images of received image data; (iii) identify one or more bins of the plurality of bins associated with a second set of datapoints corresponding to the second plurality of objects; (iv) for each bin, compute a respective sampling weight based upon the first set of datapoints and the second set of datapoints; (v) identify a bin having the respective sampling weight that satisfies a particular criterion; and (vi) add one or more objects of the second plurality of objects having corresponding datapoints of the second set of datapoints to the bin and add the image to the current training dataset.

Abstract: Aspects of this technical solution can include identifying, by a processor coupled to non-transitory memory, a plurality of bounding boxes for one or more objects depicted in each image of a sequence of images captured during operation of an autonomous vehicle, allocating, by the processor and based on corresponding positions of the bounding boxes in each image and corresponding time stamps, one or more of the bounding boxes to one or more tracking identifiers each indicating trajectories of corresponding objects, generating, by the processor and based on the time stamps and the bounding boxes allocated to each of the tracking identifiers, one or more tracking images for each of the tracking identifiers, each of the tracking images including visual indications of the time stamps, and training, by the processor and based on the tracking images, an artificial intelligence model to output an indication of a type of trajectory.

Abstract: An autonomous vehicle comprises one or more processors. The processors can be configured to receive, from a sensor of the autonomous vehicle, an image of an environment outside of the autonomous vehicle. The processors can detect potential unknown objects based on the image. The processors can compare the detection based on the image to a set of data points of a LiDAR scan to determine if there are unknown objects on a roadway.

Abstract: Aspects of this technical solution can include identifying, by a processor coupled to non-transitory memory, a plurality of bounding boxes for one or more objects depicted in each image of a sequence of images captured during operation of an autonomous vehicle, allocating, by the processor and based on corresponding positions of the bounding boxes in each image and corresponding time stamps, one or more of the bounding boxes to one or more tracking identifiers each indicating trajectories of corresponding objects, generating, by the processor and based on the time stamps and the bounding boxes allocated to each of the tracking identifiers, one or more tracking images for each of the tracking identifiers, each of the tracking images including visual indications of the time stamps, and training, by the processor and based on the tracking images, an artificial intelligence model to output an indication of a type of trajectory.

Abstract: A method of video encoding is described. The method includes segmenting original video data to obtain an original video segment including multiple video images. Video content analysis is performed on the original video segment to obtain a video image processing parameter corresponding to the original video segment. Image processing is performed on a video image in the multiple video images in the original video segment based on the video image processing parameter to obtain a processed video segment. An encoding parameter of the processed video segment can be determined based on image feature data of the processed video segment. The processed video segment can be encoded based on the encoding parameter to obtain an encoded video segment.

Abstract: A method for automatically digitally annotating a three-dimensional object in a scene includes capturing an image of the scene using a camera, capturing a point cloud representing the scene using a LiDAR system, and determining a two-dimensional boundary of the three-dimensional object contained in the image. The method further includes determining a subset of points of the point cloud contained within the two-dimensional boundary and assigning a unique identifier to the subset of points contained within the two-dimensional boundary.

Abstract: This application relates to a video encoding method, computer-readable storage medium, and computer device. The method includes: obtaining video data, the video data including a plurality of original video images; processing the video data according to a sequence generation rule to obtain a plurality of independent subsequence video image groups, each of the subsequence video image groups including subsequence images corresponding to the sequence generation rule; obtaining encoding manners corresponding to the subsequence video image groups; and independently encoding the corresponding subsequence video image groups according to the encoding manners to obtain subsequence video encoded data corresponding to each of the subsequence video image groups.

Abstract: Vehicle perception techniques include applying a 3D DNN to a set of inputs to generate 3D detection results including a set of 3D objects, transforming the set of 3D objects onto a set of images as a first set of 2D bounding boxes, applying a 2D DNN to the set of images to generate 2D detection results including a second set of 2D bounding boxes, calculating mean average precision (mAP) values based on a comparison between the first and second sets of 2D bounding boxes, identifying a set or corner cases based on the calculated mAP values, and re-training or updating the 3D DNN using the identified set of corner cases, wherein a performance of the 3D DNN is thereby increased without the use of expensive additional manually and/or automatically annotated training datasets.

Abstract: Aspects of this technical solution can include obtaining, by one or more processors coupled with non-transitory memory, a first plurality of images each including corresponding first time stamps, the plurality of images corresponding to video data of a physical environment, extracting, by the one or more processors and from among the first plurality of images, a second plurality of images each having corresponding second time stamps and each having corresponding features indicating an object in the physical environment, and training, by the one or more processors and with input including the features and the second time stamps corresponding to the features, a machine learning model to generate an output indicating a pattern of movement of one or more objects corresponding to the features and the second time stamps corresponding to the features.

Abstract: An autonomous vehicle comprises one or more processors. The processors can be configured to receive, from a sensor of the autonomous vehicle, an image of an environment outside of the autonomous vehicle. The processors can detect potential unknown objects based on the image. The processors can compare the detection based on the image to a set of data points of a LiDAR scan to determine if there are unknown objects on a roadway.

Abstract: Vehicle perception techniques include obtaining a training dataset represented by N training histograms, in an image feature space, corresponding to N training images, K-means clustering the N training histograms to determine K clusters with respective K respective cluster centers, wherein K and N are integers greater than or equal to one and K is less than or equal to N, comparing the N training histograms to their respective K cluster centers to determine maximum in-class distances for each of K clusters, applying a deep neural network (DNN) to input images of the set of inputs to output detected/classified objects with respective confidence scores, obtaining adjusted confidence scores by adjusting the confidence scores output by the DNN based on distance ratios of (i) minimal distances of input histograms representing the input images to the K cluster centers and (ii) the respective maximum in-class.

Abstract: Aspects of this technical solution can include identifying, by a processor coupled to non-transitory memory, a plurality of bounding boxes for one or more objects depicted in each image of a sequence of images captured during operation of an autonomous vehicle, allocating, by the processor and based on corresponding positions of the bounding boxes in each image and corresponding time stamps, one or more of the bounding boxes to one or more tracking identifiers each indicating trajectories of corresponding objects, generating, by the processor and based on the time stamps and the bounding boxes allocated to each of the tracking identifiers, one or more tracking images for each of the tracking identifiers, each of the tracking images including visual indications of the time stamps, and training, by the processor and based on the tracking images, an artificial intelligence model to output an indication of a type of trajectory.

Abstract: A superconducting machine includes at least one superconducting coil and a coil support structure arranged with the at least one superconducting coil. The coil support structure includes at least one composite component affixed to the at least one superconducting coil and an interface component in frictional contact with the at least one composite component so as to reduce a likelihood of quench of the at least one superconducting coil.

Abstract: Systems and methods for training artificial intelligence models based on sequences of image data are disclosed. The techniques described herein include generating, using an artificial intelligence model, a respective classification and a respective bounding box for an object depicted in each image of a sequence of images captured during operation of an autonomous vehicle; tracking the object in the sequence of images based on the respective bounding box of each image of the sequence of images and a tracking identifier corresponding to the object; determining a correction to the respective classification of an image of the sequence of images responsive to tracking the object in the sequence of images; and training the artificial intelligence model based on the correction.

Abstract: An object detection and classification verification system for a vehicle includes a projection system configured to project a three-dimensional (3D) scene pre-captured at a known distance and comprising at least one known object onto a surface in front of the vehicle a controller configured to verify a performance of an object detection and classification routine by performing the object detection and classification routine on the projected 3D scene to generate a set of results, comparing the set of results to a set of expected results associated with the projected 3D scene, and based on the comparing, determining whether the performance of the object detection and classification routine satisfies a predetermined threshold metric.

Abstract: Vehicle perception techniques include applying a 3D DNN to a set of inputs to generate 3D detection results including a set of 3D objects, transforming the set of 3D objects onto a set of images as a first set of 2D bounding boxes, applying a 2D DNN to the set of images to generate 2D detection results including a second set of 2D bounding boxes, calculating mean average precision (mAP) values based on a comparison between the first and second sets of 2D bounding boxes, identifying a set or corner cases based on the calculated mAP values, and re-training or updating the 3D DNN using the identified set of corner cases, wherein a performance of the 3D DNN is thereby increased without the use of expensive additional manually and/or automatically annotated training datasets.

Abstract: Vehicle perception techniques include obtaining a training dataset represented by N training histograms, in an image feature space, corresponding to N training images, K-means clustering the N training histograms to determine K clusters with respective K respective cluster centers, wherein K and N are integers greater than or equal to one and K is less than or equal to N, comparing the N training histograms to their respective K cluster centers to determine maximum in-class distances for each of K clusters, applying a deep neural network (DNN) to input images of the set of inputs to output detected/classified objects with respective confidence scores, obtaining adjusted confidence scores by adjusting the confidence scores output by the DNN based on distance ratios of (i) minimal distances of input histograms representing the input images to the K cluster centers and (ii) the respective maximum in-class.

Abstract: A superconducting machine includes at least one superconducting coil and a coil support structure arranged with the at least one superconducting coil. The coil support structure includes at least one composite component affixed to the at least one superconducting coil and an interface component in frictional contact with the at least one composite component so as to reduce a likelihood of quench of the at least one superconducting coil.