Abstract: A method, apparatus, non-transitory computer readable medium, and system for panoptic segmentation are described. Embodiments may generate a feature pyramid for an input image, wherein the feature pyramid comprises a plurality of feature maps at different resolution levels, apply an attention module to the feature pyramid to produce an enhanced feature map, combine the enhanced feature map with each of the plurality of feature maps to produce an enhanced feature pyramid, generate semantic information for the input image based on the enhanced feature pyramid, generate a plurality of object regions based on the enhanced feature pyramid, generate instance information for each of the plurality of object regions, and generate panoptic segmentation information for the input image based on the semantic information and the instance information for each of the plurality of object regions.

Abstract: The present invention relates to the determination of damage to portions of a vehicle. More particularly, the present invention relates to determining whether each part of a vehicle should be classified as damaged or undamaged and optionally the severity of the damage to each part of the damaged vehicle including segmenting the input images. Aspects and/or embodiments seek to provide a computer-implemented method for determining damage states of each part of a damaged vehicle, indicating whether each part of the vehicle is damaged or undamaged and optionally the severity of the damage to each part of the damaged vehicle, using images of the damage to the vehicle and trained models to assess the damage indicated in the images of the damaged vehicle, including performing segmentation of the images to create richer input data.

Abstract: A method and an apparatus for classifying a video are provided. The method may include: acquiring a to-be-classified video; extracting a set of multimodal features of the to-be-classified video; inputting the set of multimodal features into a post-fusion model corresponding to each modal respectively, to obtain multimodal category information of the to-be-classified video; and fusing the multimodal category information of the to-be-classified video, to obtain category information of the to-be-classified video. This embodiment improves the accuracy of video classification.

Abstract: The disclosure relates to method and system for training an artificial neural network (ANN) based image classifier using class-specific relevant features. The method includes receiving the ANN based image classifier, training image dataset, and various features of the training image dataset. The method further includes determining a relative relevance value of each of the features corresponding to each of the classes based on the ANN based image classifier, segregating co-occurring features from the features for each of the classes based on the training image dataset and the ANN based image classifier, identifying an imbalance in the class-specific relevant features for each of the classes based on the relative relevance value of each of the features corresponding to each of the classes, and updating the ANN based image classifier based on the imbalance in the class-specific relevant features and the co-occurring features for each of the classes.

Abstract: The present invention relates to the determination of repair operations for a damaged vehicle. More particularly, the present invention relates to determining repair operations, for example whether to repair or replace parts of a damaged vehicle and associated labour time required, for a damaged vehicle using images of the damage to the vehicle. Aspects and/or embodiments seek to provide a computer-implemented method for determining repair operations that are required to repair a damaged vehicle, using images of the damage to the damaged vehicle.

Abstract: Some embodiments of devices, systems, and methods generate a respective connection mask at each defect detection location in a binary defect map, wherein each detection location in the binary defect map has a value that indicates that the detection location either has a defect or, alternatively, does not have a defect; generate one or more clusters of defect detection locations in the binary defect map based on each defect location's respective connection mask and on the binary defect map; generate respective bounding boxes for the one or more clusters; and recursively perform the following: checking respective sizes of the bounding boxes, and splitting any clusters of the one or more clusters that have respective bounding boxes with sizes that exceed one or more thresholds, thereby generating additional clusters of defect locations.

Abstract: A neural network is configured to process image data captured by a vehicle-mounted camera. The neural network includes common processing layers (a trunk) and separate, parallelizable processing layers (branches). An object detection branch of the neural network is trained to detect objects that may be visible from the vehicle-mounted camera, such as cars, trucks, and traffic signs. A curve determination branch is trained to detect and parameterize salient curves, such as lane boundaries and road boundaries. The curve determination branch itself is configured with a trunk and branch architecture, having both common and separate processing layers. A first branch computes a likelihood that a curve is present in a given location of the image data and a second branch further localizes the curve within the given location if such a curve is present. Training of the different branches of the neural network may be decoupled.

Abstract: The discussion relates to context-aware environments. One example can include inwardly-facing cameras positioned around a periphery of an environment that defines a volume. The example can also include sensors positioned relative to the volume and configured to communicate with a user device in the volume. The example can also include an ambient perception component configured to track user locations in the volume and to detect user gestures relative to objects in the volume, and responsive to receiving a query from the user's device, to supplement the query with information derived from the objects.

Abstract: An object re-identifier. For each of a plurality of frames of a video, a quality of the frame is assessed and a confidence that a previously-recognized object is present in the frame is determined. The determined confidence for the frame is weighted based on the assessed quality of the frame such that frames with higher relative quality are weighted more heavily than frames with lower relative quality. An overall confidence that the previously-recognized object is present in the video is assessed based on the weighted determined confidences.

Abstract: Disclosed is a method for providing user-customized food information. In the method for providing user-customized food information, when an input of a user is received, the user is classified, history related to the classified user is retrieved from among previously stored food entry/exit history, and information on preferred food of the classified user on the basis of the retrieved entry/exit history, thereby easily providing intuitive information on preferred food having a high usage frequency for each user. The device for providing user-customized food information of the present disclosure is linked to an artificial intelligence module, a robot, an augmented reality (AR) device, a virtual reality (VR) device, a device related to a 5G service, and the like. Can be.

Abstract: An electronic system is disclosed comprising of a drill tool which is mounted on a mechanical system with either a position and angular sensor assisted guidance of the camera itself around the drill tool in a guided trajectory as in a robotic arm or mobile device on the field with the drill tool pulled out from the wellbore or when the tool can be imaged while in operation; multiple fixed cameras on a test or working rig coupled with rotary control of the drill tool, and a controller capable of controlling the drill string motor and imaging. The system automatically profiles the degraded/worn out drill tool, when the drill string-drill tool is visually available or pulled out, by executing a host of machine vision engineering based and machine intelligence/data analytics engineering algorithms, where the algorithms can calculate and provide detailed current/predictive and prescriptive analysis of the drilling tool.

Abstract: Non-contact methods and systems are disclosed for estimating an in-bed human pose. The method includes the steps of: (a) capturing thermal imaging data of a human subject lying on a bed using a long wavelength infrared camera positioned above the human subject; (b) transmitting the thermal imaging data to a computer system; and (c) processing the thermal imaging data by the computer system using a model to estimate the pose of the human subject, the model comprising a machine learning inference model trained on a training dataset of a plurality of in-bed human poses.

Abstract: A method for object detection includes: obtaining a plurality of to-be-determined targets in a to-be-detected image; determining confidences of the plurality of to-be-determined targets separately belonging to at least one category, determining categories of the plurality of to-be-determined targets according to the confidences, and determining position offset values corresponding to the respective categories of the plurality of to-be-determined targets; using the position offset values corresponding to the respective categories of the plurality of to-be-determined targets as position offset values of the plurality of to-be-determined targets; and determining position information and a category of at least one to-be-determined target in the to-be-detected image according to the categories of the plurality of to-be-determined targets, the position offset values of the plurality of to-be-determined targets, and the confidences of the plurality of to-be-determined targets belonging to the categories thereof.

Abstract: A vehicle exterior environment recognition apparatus includes a position derivation unit, a grouping unit, and a speed derivation unit. The position derivation unit is configured to derive three-dimensional positions of respective blocks from an image received from an imaging device. The grouping unit is configured to group the respective blocks and to identify a three-dimensional object. The speed derivation unit is configured to make a determination as to whether a side of the three-dimensional object satisfies a determination condition that the side of the three-dimensional object is included in an end area situated on an end side of the image in a horizontal direction. The speed derivation unit is configured to derive the speed of the three-dimensional object in accordance with an end of the side of the three-dimensional object in a depth direction when the determination is made that the side of the three-dimensional object satisfies the determination condition.

Abstract: The present disclosure relates to an artificial intelligence (AI) system for simulating functions of a human brain such as cognition and decision-making by using machine learning algorithms such as deep learning, and applications thereof. In particular, the present disclosure provides a method of recognizing an object by using an AI system and its application, including: extracting pieces of first feature information respectively regarding a plurality of images, each image including an object; generating at least one piece of second feature information representing a correlation between the plurality of images by combining together the extracted pieces of first feature information respectively regarding the plurality of images; and recognizing, based on the at least one piece of second feature information, the object included in each of the plurality of images by using a pre-generated learning network model.

Abstract: The disclosure relates to the technical field of data processing. Disclosed are a pedestrian flow funnel generation method and apparatus, a storage medium and an electronic device. The method can comprise: acquiring a current frame of image, and tracking and updating, according to a multi-target tracking algorithm, a head and shoulder area in a tracking sequence set in the current frame of image; acquiring, according to a head and shoulder recognition model, a head and shoulder area in the current frame of image, and updating, according to the head and shoulder area in the current frame of image, the tracking sequence set; analysing a motion track of each head and shoulder area in the tracking sequence set so as to count pedestrians, and when the current frame of image is the last frame of image, generating a pedestrian flow funnel based on a counting result of pedestrians.

Abstract: Systems and methods are provided for identifying and representing a traffic signal device. The method includes determining a location and orientation of the vehicle and receiving a real world image. The method further includes analyzing information about the vehicle's location and environment and using this information and the vehicle's orientation to generate a raster image illustrating an approximation of a view of the real world image, including one or more traffic signal devices. Additionally, the method includes providing the real world image and the raster image as inputs to a neural network to classify a traffic signal device in the real world image as the primary traffic signal device and determine a set of coordinates indicating a location of the primary traffic signal device, generating a classified real world image which includes a bounding box indicating the set of coordinates, and receiving the classified real world image.

Abstract: According to the present disclosure, an artificial intelligence (AI) system and a method of providing content according to an application of the AI system are provided. The method includes: obtaining one or more images included in the content; generating additional content for guiding user information, the additional content corresponding to the one or more images, based on feature information extracted from the one or more images; when receiving a request for reproducing the content, synchronizing the generated additional content with the one or more images; and reproducing the content and the additional content, according to a result of the synchronizing.

Abstract: A device includes image generation circuitry and convolutional-neural-network circuitry. The image generation circuitry, in operation, generates a digital image representation of a wafer defect map (WDM). The convolutional-neural-network circuitry, in operation, generates a defect classification associated with the WDM based on: the digital image representation of the WDM and a data-driven model associating WDM images with classes of a defined set of classes of wafer defects and generated using a training data set augmented based on defect pattern orientation types associated with training images.

Abstract: A template determining apparatus including an attribute distribution determination unit configured to determine a distribution of a specific attribute in a plurality of images; and a template determination unit configured to adaptatively determine a template set from the plurality of images according to the determined distribution of the specific attribute of the plurality of images. Where the determined template set will be used for image normalization.