Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating a final classification output for an image of eye tissue. The image is provided as input to each of one or more segmentation neural networks to obtain one or more segmentation maps of the eye tissue in the image. A respective classification input is generated from each of the segmentation maps. For each of the segmentation maps, the classification input for the segmentation map is provided as input to each of one or more classification neural networks to obtain, for each segmentation map, a respective classification output from each classification neural network. A final classification output for the image is generated from the respective classification outputs for each of the segmentation maps.

Abstract: The system of the present disclosure comprises: an acquisition means for acquiring a video image of an online session between a first user and a second user; a face recognition means for recognizing at least a face image of the first user and the second user included in the video image for each predetermined frame; a voice recognition means for recognizing at least the voice of the subject included in the video image; an evaluation means for calculating an evaluation value from a plurality of viewpoints based on both the recognized face image and the voice; and a determination means for determining the degree of match of the second user to the first user based on the evaluation values.

Abstract: Techniques related to automatically segmenting video frames into per pixel fidelity object of interest and background regions are discussed. Such techniques include applying tessellation to a video frame to generate feature frames corresponding to the video frame and applying a segmentation network implementing context aware skip connections to an input volume including the feature frames and a context feature volume corresponding to the video frame to generate a segmentation for the video frame.

Abstract: Embodiments of the present disclosure include a method that obtains a digital image. The method includes extracting a word block from the digital image. The method includes processing the word block by evaluating a value of the word block against a dictionary. The method includes outputting a prediction equal to a common word in the dictionary when a confidence factor is greater than a predetermined threshold. The method includes processing the word block and assigning a descriptor to the word block corresponding to a property of the word block. The method includes processing the word block using the descriptor to prioritize evaluation of the word block. The method includes concatenating a first output and a second output. The method includes predicting a value of the word block.

Abstract: A process for reducing time of transmission for single-band, multiple-band or hyperspectral imagery using Machine Learning based compression is disclosed. The process uses Machine Learning to compress single-band, multiple-band and hyperspectral imagery, thereby decreasing the needed bandwidth and storage-capacity requirements for efficient transmission and data storage. The reduced file size for transmission accelerate the communications and reduces the transmission time. This enhances communications systems where there is a greater need for on or near real-time transmission, such as mission critical applications in national security, aerospace and natural resources.

Abstract: A device, method, and non-transitory computer readable medium are described. The method includes receiving a dataset including hand written Arabic words and hand written Arabic alphabets from one or more users. The method further includes removing whitespace around alphabets in the hand written Arabic words and the hand written Arabic alphabets in the dataset. The method further includes splitting the dataset into a training set, a validation set, and a test set. The method further includes classifying one or more user datasets from the training set, the validation set, and the test set. The method further includes identifying the target user from the one or more user datasets. The identification of the target user includes a verification accuracy of the hand written Arabic words being larger than a verification accuracy threshold value.

Abstract: Disclosed is a data recognition model construction apparatus. The data recognition model construction apparatus includes a video inputter configured to receive a video, an image composition unit configured to, based on a common area included in each of a plurality of images that form at least a portion of the video, generate a composition image by overlaying at least a portion of the plurality of images, a learning data inputter configured to receive the generated composition image, a model learning unit configured to make a data recognition model learn using the generated composition image, and a model storage configured to store the learnt data recognition model.

Abstract: To implement a video processing device, a video processing method, and a video processing program capable of estimating a movement vector from a video content and providing processing information based on the movement vector to a haptic device or other force sense presentation devices, a video processing device according to the present disclosure includes a scene identification unit to estimate scene class information that is information identifying a scene class for a video content and a plurality of movement information estimation units to estimate a movement vector from the video content. One movement information estimation unit that is selected from the plurality of the movement information estimation units in response to the scene class identified by the scene class information estimates the movement vector.

Abstract: This application relates to a method and a system for building machine learning or deep learning data sets for automatically recognizing labels on items. The system may include an optical scanner configured to capture an item including one or more labels provided thereon, the item captured a plurality of times at different positions with respect to the optical scanner. The system may further include a robotic arm on which the item is disposed, the robotic arm configured to rotate the item horizontally and/or vertically such that the one or more labels of the item are captured by the optical scanner at different positions with respect to the optical scanner. The system may include a database configured to store the captured images.

Abstract: A method for adding biometric authentication training data into databases performed by a biometric authentication data classification device includes: extracting first biometric characteristic information from at least one candidate biometric training data for biometric authentication using an artificial neural network model; calculating an overall similarity between the first biometric characteristic information and second biometric characteristic information extracted from a performance test database of which a biometric authentication performance is lower than a threshold level, the performance test database being selected among performance test databases for the biometric authentication; and adding the at least one candidate biometric training data into one of the biometric authentication training database and the performance test database based on the calculated overall similarity.

Abstract: In a method of color decomposition, inter-color images indicating similarity between color sensitivities are generated based on color images. Conversion coefficients of the color images and the inter-color images with respect to a white image are determined. A pseudo-white image corresponding to the color images and the inter-color images is generated using the conversion coefficients the pseudo-white image similar to a real white image is generated using the inter-color images indicating similarity between color sensitivities. Deep learning of the artificial neural network is performed efficiently using the color images and the pseudo-white image and the demosaiced images of high quality are generated using the trained artificial neural network that is trained.

Abstract: A vehicle device setting method including: capturing, by an image sensing unit, a first image frame; recognizing a user ID according to the first image frame; showing ID information of the recognized user ID on a screen or by a speaker; capturing a second image frame; generating a confirm signal when a first user expression is recognized by calculating an expression feature in the second image frame and comparing the recognized expression feature with stored expression data associated with a predetermined user expression to confirm whether the recognized user ID is correct or not according to the second image frame captured after the ID information is shown; controlling an electronic device according to the confirm signal; and entering a data update mode instructed by the user and updating setting information of the electronic device by current electronic device setting according to a saving signal generated by confirming a second user expression in a third image frame captured after the user ID is confirmed

Abstract: Examples are disclosed herein that relate to automatically calibrating cameras based on human detection. One example provides a computing system comprising instructions executable to receive image data comprising depth image data and two-dimensional image data of a space from a camera, detect a person in the space via the image data, determine a skeletal representation for the person via the image data, determine over a period of time a plurality of locations at which a reference point of the skeletal representation is on a ground area in the image data, determine a ground plane of the three-dimensional representation based upon the plurality of locations at which the reference point of the skeletal representation is on the ground area in the image data, and track a location of an object within the space relative to the ground plane.

Abstract: Described is a system and method for enabling visual search for information. With each selection of an object included in an image, additional images that include visually similar objects are determined and presented to the user.

Abstract: A system and method of generating a player tracking prediction are described herein. A computing system retrieves a broadcast video feed for a sporting event. The computing system segments the broadcast video feed into a unified view. The computing system generates a plurality of data sets based on the plurality of trackable frames. The computing system calibrates a camera associated with each trackable frame based on the body pose information. The computing system generates a plurality of sets of short tracklets based on the plurality of trackable frames and the body pose information. The computing system connects each set of short tracklets by generating a motion field vector for each player in the plurality of trackable frames. The computing system predicts a future motion of a player based on the player's motion field vector using a neural network.

Abstract: This application discloses a semantic learning-based down-sampling method of point cloud data of an aircraft, including: (S1) constructing a multi-input encoder based on feature learning according to point cloud semantic learning principle; inputting the point cloud data of the aircraft and feature point data into the multi-input encoder for feature fusion followed by decoding using a decoder the multi-input feature fused data to obtain to-be-measured data; (S2) constructing and training a point cloud feature weight calculation network based on semantic learning to acquire a feature weight of each point in the to-be-measured data; and (S3) performing spatial weighted sampling on the feature weight of each point in the to-be-measured data followed by down-sampling based on Gaussian distribution-based spatial sampling principle.

Abstract: An apparatus for performing artificial intelligence (AI) encoding on an image includes: a memory storing one or more instructions; and a processor configured to execute the one or more instructions stored in the memory to: determine a resolution of an original image; when the resolution of the original image is higher than a predetermined value, obtain a first image by performing AI downscaling on the original image via a downscaling deep neural network (DNN); when the resolution of the original image is lower than or equal to the predetermined value, obtain a first image by performing AI one-to-one preprocessing on the original image via a one-to-one preprocessing DNN for upscaling; generate image data by performing first encoding on the first image; and transmit the image data and AI data including information related to the AI downscaling or information related to the AI one-to-one preprocessing.

Abstract: An artificial intelligence (AI) decoding method including obtaining image data generated from performing first encoding on a first image and AI data related to AI down-scaling of at least one original image related to the first image; obtaining a second image corresponding to the first image by performing first decoding on the image data; obtaining, based on the AI data, deep neural network (DNN) setting information for performing AI up-scaling of the second image; and generating a third image by performing the AI up-scaling on the second image via an up-scaling DNN operating according to the obtained DNN setting information. The DNN setting information is DNN information updated for performing the AI up-scaling of at least one second image via joint training of the up-scaling DNN and a down-scaling DNN used for the AI down-scaling.

Abstract: The present disclosure includes a method, apparatus and system for generating a neural network and a non-transitory computer readable storage medium storing instructions. The method comprises: recognizing at least an attribute of an object in a sample image according to a feature extracted from the sample image, using the neural network; determining a loss function value at least according to a margin value determined based on a semantic relationship between attributes, wherein the semantic relationship is obtained from a predefined table at least according to a real attribute and the recognized attribute of the object, wherein the predefined table is composed of the attributes and the semantic relationship between the attributes; updating a parameter in the neural network according to the determined loss function value. When using the neural network generated according the present disclosure, the accuracy of object attribute recognition can be improved.

Abstract: Techniques for training an optical character recognition (OCR) model to detect and recognize text in images for robotic process automation (RPA) are disclosed. A text detection model and a text recognition model may be trained separately and then combined to produce the OCR model. Synthetic data and a smaller amount of real, human-labeled data may be used for training to increase the speed and accuracy with which the OCR text detection model and the text recognition model can be trained. After the OCR model has been trained, a workflow may be generated that includes an activity calling the OCR model, and a robot implementing the workflow may be generated and deployed.