Abstract: A method for training generative network, a method for generating near-infrared image and device. The method includes: obtaining a training sample set, in which the set includes near-infrared image samples and visible-light image samples; obtaining an adversarial network to be trained, in which the generative network of the adversarial network is configured to generate each near-infrared image according to an input visible-light image, the discrimination network of the adversarial network is configured to determine whether an input image is real or generated; constructing a first objective function according to a first distance between each generated near-infrared image and the corresponding near-infrared image sample in an image space and a second distance between each generated near-infrared image and the corresponding near-infrared image sample in a feature space; performing an adversarial training on the adversarial network with the set based on optimizing a value of the first objective function.

Abstract: Driver behavior risk assessment and pedestrian awareness may include an receiving an input stream of images of an environment including one or more objects within the environment, estimating an intention of an ego vehicle based on the input stream of images and a temporal recurrent network (TRN), generating a scene representation based on the input stream of images and a graph neural network (GNN), generating a prediction of a situation based on the scene representation and the intention of the ego vehicle, and generating an influenced or non-influenced action determination based on the prediction of the situation and the scene representation.

Abstract: Embodiments are disclosed for deblurring motion in video. A method of deblurring motion in video can include receiving an input frame from a digital video, extracting a plurality of features of the input frame using an encoder network, determining, using a neural network, a plurality of spatial alignment kernels and a plurality of deblur kernels each corresponding to a feature of the input frame, wherein the plurality of spatial alignment kernels include different sizes of spatial alignment kernels and wherein the plurality of deblur kernels include different sizes of deblur kernels, generating, by the neural network, a plurality of output features for the input frame using the plurality of spatial alignment kernels and the plurality of deblur kernels, and generating a deblurred output frame from the plurality of output features using a decoder network.

Abstract: Systems and techniques are provided for performing holistic video understanding. For example a process can include obtaining a first video and determining, using a machine learning model decision engine, a first machine learning model from a set of machine learning models to use for processing at least a portion of the first video. The first machine learning model can be determined based on one or more characteristics of at least the portion of the first video. The process can include processing at least the portion of the first video using the first machine learning model.

Abstract: A controlling method of an electronic apparatus may include: obtaining image data and metadata regarding the image data, the image data comprising a first image frame and a second image frame that is subsequent to the first image frame; obtaining information regarding a region of interest of the first image frame by inputting the first image frame to a first neural network model; obtaining a similarity between the first image frame and the second image frame based on motion vector information included in the metadata; and detecting whether there is a manipulated area in the second image frame based on the similarity.

Abstract: A method for generating an output image may include obtaining an input image having a first exposure value, generating a plurality of levels that are each respectively associated with a respective representation of the input image, based on the input image, generating the output image having a second exposure value, based on a deep neural network (DNN) including a set of sub-networks and the plurality of levels, and providing the output image.

Abstract: The technology disclosed relates to constructing a convolutional neural network-based classifier for variant classification. In particular, it relates to training a convolutional neural network-based classifier on training data using a backpropagation-based gradient update technique that progressively match outputs of the convolutional neural network-based classifier with corresponding ground truth labels. The convolutional neural network-based classifier comprises groups of residual blocks, each group of residual blocks is parameterized by a number of convolution filters in the residual blocks, a convolution window size of the residual blocks, and an atrous convolution rate of the residual blocks, the size of convolution window varies between groups of residual blocks, the atrous convolution rate varies between groups of residual blocks. The training data includes benign training examples and pathogenic training examples of translated sequence pairs generated from benign variants and pathogenic variants.

Abstract: A leg (205) detection system comprising: a robotic arm (200) comprising a gripping portion (208) for holding a teat cup (203, 210) for attaching to a teat (1102, 1104, 1106, 1108, 203S, 203) of a dairy livestock (200, 202, 203); an imaging system coupled to the robotic arm (200) and configured to capture a first three-dimensional (3D) image (138, 2400, 2500) of a rearview of the dairy livestock (200, 202, 203) in a stall (402), the imaging system comprising a 3D camera (136, 138) or a laser (132), wherein each pixel of the first 3D image (138, 2400, 2500) is associated with a depth value; one or more memory (104) devices configured to store a reference (3D) 3D image (138, 2400, 2500) of the stall (402) without any dairy livestock (200, 202, 203); and a processor (102) communicatively coupled to the imaging system and the one or more memory (104) devices, the processor (102) configured to: access the first 3D image (138, 2400, 2500) and the reference (3D) 3D image (138, 2400, 2500); subtract the first 3D image

Abstract: Techniques for determining a classification probability of an object in an environment are discussed herein. Techniques may include analyzing sensor data associated with an environment from a perspective, such as a top-down perspective, using multi-channel data. From this perspective, techniques may determine channels of multi-channel input data and additional feature data. Channels corresponding to spatial features may be included in the multi-channel input data and data corresponding to non-spatial features may be included in the additional feature data. The multi-channel input data may be input to a first portion of a machine-learned (ML) model, and the additional feature data may be concatenated with intermediate output data from the first portion of the ML model, and input into a second portion of the ML model for subsequent processing and to determine the classification probabilities. Additionally, techniques may be performed on a multi-resolution voxel space representing the environment.

Abstract: A non-transitory computer readable medium (107, 127) stores instructions executable by at least one electronic processor (101, 113) to perform a component co-replacement recommendation method (200).

Abstract: Aspects described herein provide a method of processing data in a machine learning model, including: receiving first domain input data; transforming the first domain input data to second domain input data via a domain transformation function; providing the second domain input data to a first layer of a machine learning model; processing the second domain input data in the first layer of the machine learning model according to a set of layer weights; and outputting second domain output data from the first layer of the machine learning model.

Abstract: Systems and methods to estimate 3D TOF scatter include acquisition of 3D TOF data, determination of 2D TOF data from the first TOF data, determination of first estimated scatter based on the second TOF data, reconstruction of a first estimated image based on the first estimated scatter and the second TOF data, determination of attenuated unscattered true coincidences based on the first estimated image, determination of second estimated scatter based on the first TOF data and the attenuated unscattered true coincidences, and reconstruction of an image of the object based on the first TOF data and the second estimated scatter.

Abstract: The systems, methods, a non-transitory computer readable mediums relate to an object selection system that accurately detects and automatically selects user-requested objects (e.g., query objects) in a digital image. For example, the object selection system builds and utilizes an object selection pipeline to determine which object detection neural network to utilize to detect a query object based on analyzing the object class of the query object. In addition, the object selection system can add, update, or replace portions of the object selection pipeline to improve overall accuracy and efficiency of automatic object selection within an image.

Abstract: A character recognition method, a character recognition apparatus, an electronic device and a computer readable storage medium are disclosed. The character recognition method includes: determining semantic information and first position information of each individual character recognized from an image; constructing a graph network according to the semantic information and the first position information of each individual character; and determining a character recognition result of the image according to a feature of each individual character calculated by the graph network.

Abstract: A traffic light classification system for a vehicle includes an image capture device to capture an image of a scene that includes a traffic light with multiple light signals, a processor, and a memory communicably coupled to the processor and storing a first neural network module including instructions that when executed by the processor cause the processor to determine, based on inputting the image into a neural network, a semantic keypoint for each light signal in the traffic light, and determine, based on each semantic keypoint, a classification state of each light signal.

Abstract: A method and system are provided for training a machine learning (ML) system for predicting delays in processing pipelines. In one embodiment, the method includes receiving labelled historical data pertaining to a pipeline, the labelled data including trigger objects initiating the pipeline and one or more processing times corresponding to one or more stages of the pipeline. The method includes identifying features associated with the trigger objects, formatting the labelled data and, randomly splitting the formatted labelled data into a full training dataset and a testing dataset. Additionally, the method includes distributing the full training dataset into several partial datasets and, in an ensemble ML system, training each of several ML subsystems using a respective partial dataset to provide a respective individual inference model predicting respective processing times, and deriving and storing an ML model for prediction of delays by aggregating the individual inference models.

Abstract: A method for visual question answering, a computer device implementing the method and a medium for storing instructions on performing the method are provided. The method includes: acquiring an input image and an input question; constructing a visual graph based on the input image, wherein the visual graph comprises a first node feature and a first edge feature; constructing a question graph based on the input question, wherein the question graph comprises a second node feature and a second edge feature; performing a multimodal fusion on the visual graph and the question graph to obtain an updated visual graph and an updated question graph; determining a question feature based on the input question; determining a fusion feature based on the updated visual graph, the updated question graph and the question feature; and generating a predicted answer for the input image and the input question.

Abstract: An online system that allows users to interact with it using expressions in natural language form includes an intent inference module allowing it to infer an intent represented by a user expression. The intent inference module has a set of possible intents, along with a small set of example natural language expressions known to represent that intent. When a user interacts with the system using a natural language expression for which the intent is not already known, the intent inference module applies a natural language inference model to compute scores indicating whether the user expression textually entails the various example natural language expressions. Based on the scores, the intent inference module determines an intent that is most applicable for the expression. If an intent cannot be determined with sufficient confidence, the intent inference module may further attempt to determine whether the various example natural language expressions textually entail the user expression.

Abstract: A system, electronic device and method for improved neural network training are provided. The improved system is adapted for tracking long range dependence in sequential data during training, and includes bootstrapping a lower bound on the mutual information (MI) over groups of variables (segments or sentences) and subsequently applying the bound to encourage high MI.

Abstract: A knowledge distillation based semantic image segmentation method includes inputting an input image to a teacher network and a student network; normalizing a first feature vector corresponding to each pixel in a feature map of a last layer of the teacher network and normalizing a second feature vector corresponding to each pixel in a feature map of a last layer of the student network; generating the first channel and space association matrix and the second channel and space association matrix based on the normalized first feature vector and the normalized second feature vector, and defining a first loss function based on a Euclidean norm value of the difference between the first channel and space association matrix and the second channel and space association matrix.