Abstract: A non-transitory computer-readable medium including a face authentication program for causing a computer device that executes face authentication using a plurality of face information registered in a storage unit with respect to a captured face image of a user to implement the following functions including a determination function of determining face information to be used for the face authentication from among the plurality of face information based on position estimation information of a user corresponding to each of the plurality of face information, which is registered in the storage unit in association with each of the plurality of face information, and a predetermined condition, and an authentication function of performing the face authentication using the determined face information with respect to the captured face image of the user.

Abstract: Presented herein are systems, methods and apparatuses for increasing reliability of face recognition in analysis of images captured by drone mounted imaging sensors, comprising: recognizing a target person in one or more iterations, each iteration comprising: identifying one or more positioning properties of the target person based on analysis of image(s) captured by imaging sensor(s) mounted on a drone operated to approach the target person, instructing the drone to adjust its position to an optimal facial image capturing position selected based on the positioning property(s), receiving facial image(s) of the target person captured by the imaging sensor(s), receiving a face classification associated with a probability score from machine learning model(s) trained to recognize the target person, and initiating another iteration in case the probability score does not exceed a certain threshold. Finally, the face classification may be outputted for use by one or more face recognition based systems.

Abstract: A method includes obtaining a two-dimensional image, obtaining a two-dimensional image annotation that indicates presence of an object in the two-dimensional image, obtaining three-dimensional sensor information, generating a top-down representation of the three-dimensional sensor information, and obtaining a top-down annotation that indicates presence of the object in the top-down representation. The method also includes determining a bottom surface of a three-dimensional cuboid based on map information, determining a position, a length, a width, and a yaw rotation of the three-dimensional cuboid based on the top-down annotation, and determining a height of the three-dimensional cuboid based on a two-dimensional image annotation, and the position, the length, the width, and the yaw rotation of the three-dimensional cuboid.

Abstract: A method and a system of deep face recognition for a fisheye image are provided. The method includes determining a category corresponding to an input image, performing an image rectification according to the category corresponding to the input image to generate a restored image, and performing a face recognition on the restored image to determine an identity corresponding to the input image. The category correlates to a radial distance corresponding to the input image.

Abstract: An image masking method is provided. The method includes: extracting an object from an image; obtaining characteristic information about the extracted object by analyzing the extracted object; determining whether the extracted object is a masking target according to an input setting value or the obtained characteristic information; and performing masking such that the obtained characteristic information is reflected on the extracted object, in response to determining that the extracted object is the masking target among a plurality of objects extracted from the input image, wherein the setting value is set by an input designating at least a partial region in the input image, and wherein in the determining whether the extracted object is the masking target, an object positioned in the at least a partial region is determined as the masking target among the extracted objects.

Abstract: Provided are systems and methods for text and audio-based real-time face reenactment. An example method includes receiving an input text and a target image, the target image including a target face; generating, based on the input text, a sequence of sets of acoustic features representing the input text; generating, based on the sequence of sets of acoustic features, a sequence of sets of mouth key points; generating, based on the sequence of sets of mouth key points, a sequence of sets of facial key points; generating, by the computing device and based on the sequence of sets of the facial key points and the target image, a sequence of frames; and generating, based on the sequence of frames, an output video. Each of the frames includes the target face modified based on at least one set of mouth key points of the sequence of sets of mouth key points.

Abstract: A face recognition method includes: detecting keypoints when receiving a first face image; acquiring a recognition score of each detectable keypoint and serial numbers of missing keypoints; acquiring a plurality of target keypoints in the plurality of detectable keypoints having a predetermined face feature association relationship with the missing keypoints when the influence score is higher than a predetermined score threshold; acquiring a target face feature template having a degree of position combination with the plurality of target keypoints greater than a predetermined combination degree threshold; and stitching the target face feature template and the plurality of target keypoints on the first face image to obtain a second face image so as to detect all the keypoints according to the second face image for performing the face recognition.

Abstract: Implementations described herein relate to methods, systems, and computer-readable media to render a foreground video. In some implementations, a method includes receiving a plurality of video frames that include depth data and color data. The method further includes downsampling the frames of the video. The method further includes, for each frame, generating an initial segmentation mask that categorizes each pixel of the frame as foreground pixel or background pixel. The method further includes determining a trimap that classifies each pixel of the frame as known background, known foreground, or unknown. The method further includes, for each pixel that is classified as unknown, calculating and storing a weight in a weight map. The method further includes performing fine segmentation to obtain a binary mask for each frame. The method further includes upsampling the plurality of frames based on the binary mask for each frame to obtain a foreground video.

Abstract: A system and method for verifying the identity of a user sending a facial image is disclosed. The system includes a mobile device interface for receiving a facial image. A facial feature extractor module receives the image and determines identifying features from the facial image. An image template database stores a template image associated with the user of the mobile device. A match processing module receives the facial image and verifies the facial image with the identity of the user by comparing the identifying features with the template image.

Abstract: Disclosed herein are a method and apparatus for active identification based on gaze path analysis. The method may include extracting the face image of a user, extracting the gaze path of the user based on the face image, verifying the identity of the user based on the gaze path, and determining whether the face image is authentic.

Abstract: Disclosed are systems, methods, and non-transitory computer-readable media for automatic image selection for online product catalogs. An image selection system gathers feature data for images of an item included in listings posted to an online marketplace. The image selection system uses the feature data as input in a machine learning model to determine probability scores indicating an estimated probability that each image is suitable to represent the item. The machine learning model is trained based on a set of training images of the item that have been labeled to indicate whether they are suitable to represent the image. The image selection system compares the probability scores and selects an image to represent the item as a stock image based on the comparison.

Abstract: Methods, systems, and articles of manufacture to autonomously select data structures are disclosed. An example apparatus includes an ordinal assigner to assign training code operations to respective first ordered values, and assign candidate data structure types to respective second ordered values, a filter generator to, for a first instruction of the training code operations, generate a Bloom filter bit vector pattern based on (a) one of the first ordered values, (b) one of the second ordered values corresponding to a first one of the candidate data structure types, and (c) a size of the first instruction, a label generator to generate a first model training input feature vector based on the Bloom filter bit vector pattern, data corresponding to the first instruction, and a performance metric of the first one of the candidate data structure types, and a neural network manager to train the data structure selection model with the first model training input feature vector.

Abstract: A system for predicting alimentary element ordering based on biological extraction, the system comprising a computing device configured to receive a biological extraction and alimentary element order chronicle of a user, retrieve an alimentary profile, identify, using the alimentary profile and a predictive machine-learning process, a predicted alimentary element and an alternative alimentary element, determine, using the predictive machine-learning process and the alimentary profile, the predicted alimentary element, select, using the predicted alimentary element, the alternative alimentary element, create a classifier, using a classification machine-learning process as a function of a plurality of alimentary element metrics, generate a plurality of related alimentary elements as a function of the classifier, rank the related alimentary elements as a function of the biological extraction, select the alternative alimentary element as a function of the ranking, and present the predicted alimentary element and

Abstract: An object of the present disclosure is to provide a system for deriving a type of a defect of a semiconductor element and a non-transitory computer-readable medium. The system receives, from the image acquisition tool, image data obtained by sequentially irradiating a plurality of patterns provided on the semiconductor wafer with a beam and extracts characteristics of the plurality of patterns sequentially irradiated with a beam from the received image data, the characteristics being included in the image data, or receives characteristics of the plurality of patterns sequentially irradiated with a beam from the image acquisition tool, the characteristics being extracted from the image data (Step 603), and derives (Step 605) a type of a defect by referring to (Step 604) related information for the characteristics of the plurality of patterns, the related information storing the characteristics of the plurality of patterns and types of defects in association with each other.

Abstract: Disclosed are systems and methods for template-based generation of personalized videos. An example method may commence with receiving video configuration data including a sequence of frame images, a sequence of face area parameters defining positions of a face area in the frame images, and a sequence of skin masks defining positions of a skin area of a part of the at least one body in the frame images. The method may continue with receiving an image of a source face. The method may further include determining color data associated with the source face. The method may include recoloring the skin area of the part of the at least one body in the frame image and inserting the image of the source face into the frame image at a position determined by face area parameters corresponding to the frame image to generate an output frame of an output video.

Abstract: A method and system for detecting facial expressions in digital images and applications therefore are disclosed. Analysis of a digital image determines whether or not a smile and/or blink is present on a person's face. Face recognition, and/or a pose or illumination condition determination, permits application of a specific, relatively small classifier cascade.

Abstract: Systems and methods are provided for determining a candidate identity of a person in video data based on a biometric motion signature corresponding to motion of a face of the person in the video data. Motion information is obtained corresponding to motion of the face in the video data. The motion information is provided as input to a neural network and the neural network generates the biometric motion signature in response. The biometric motion signature is compared with a plurality of biometric signatures stored in data storage. The candidate identity of the person is determined as a result of a correspondence between the biometric motion signature and a stored biometric signature of the plurality of biometric signatures.

Abstract: A digital image is stored on a server. One or more areas of interest are determined within the digital image. One or more sets of features for each of the one or more areas of interest within the digital image are extracted. The digital image is parsed into a set of image layers, wherein a subset of image layers is associated with a first set of features of the one or more sets of features. A request to download the image from the server is received from a client. The subset of image layers is sent to the client. In response to sending the subset, the remainder of the set of images is sent.

Abstract: The present disclosure relates to methods and systems that facilitate determination of a pose of a vehicle based on various combinations of map data and sensor data received from light detection and ranging (LIDAR) devices and/or camera devices. An example method includes receiving point cloud data from a (LIDAR) device and transforming the point cloud data to provide a top-down image. The method also includes comparing the top-down image to a reference image and determining, based on the comparison, a yaw error. An alternative method includes receiving camera image data from a camera and transforming the camera image data to provide a top-down image. The method also includes comparing the top-down image to a reference image and determining, based on the comparison, a yaw error.

Abstract: A video input interface receives input video data in a normal state. A control input interface receives character data for On Screen Display (OSD) in a setup state. An encoder encodes the character data and stores encoded compressed data in a memory in the setup state. A decoder receives an instruction signal designating the character data to be displayed, reads and decodes one piece of compressed data corresponding to the instruction signal from the memory, and reproduces the original character data in the normal state. A multiplexer superimposes character data on frame data and outputs the data.