Abstract: An example embodiment includes: a determination unit that, based on an image including an eye of a recognition subject, determines whether or not a colored contact lens is worn; and a processing unit that, when it is determined by the determination unit that the colored contact lens is worn, performs a process of improving accuracy of iris matching on the recognition subject.

Abstract: Embodiments of the present invention disclose a method, computer program product, and system for avoiding augmented reality blindness during an augmented reality session via an augmented reality device used by a user. In the method, computer program product, and system a set of visual representations within the augmented reality session is detected. A set of labeling objects is applied to the set of visual representations within the augmented reality session. A subset of labeling objects and associated visual representation within a display view of the augmented reality device are recorded in a history table. A blindness quotient based on the history table of recorded labeled objects and associated visual representation is calculated. In response to the calculated blindness quotient exceeding a threshold, the set of labeling objects within the display view of the augmented reality device are modified.

Abstract: The disclosed computer-implemented method may include receiving, from devices in an environment, real-time data associated with the environment. The method may also include determining, from the real-time data, current mapping and object data. The current mapping data may include coordinate data for the environment and the current object data may include both state data and relationship data for objects in the environment. The method may also include determining mapping deltas between the current mapping data and baseline map data and determining object deltas between the current object data and an event graph. The event graph may include prior state data and prior relationship data for objects. The method may also include updating the baseline map data and the event graph based on the deltas and sending updated baseline map data and event graph data to the devices. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: The disclosure herein describes training a machine learning model to recognize a real-world object based on generated virtual scene variations associated with a model of the real-world object. A digitized three-dimensional (3D) model representing the real-world object is obtained and a virtual scene is built around the 3D model. A plurality of virtual scene variations is generated by varying one or more characteristics. Each virtual scene variation is generated to include a label identifying the 3D model in the virtual scene variation. A machine learning model may be trained based on the plurality of virtual scene variations. The use of generated digital assets to train the machine learning model greatly decreases the time and cost requirements of creating training assets and provides training quality benefits based on the quantity and quality of variations that may be generated, as well as the completeness of information included in each generated digital asset.

Abstract: A method and an apparatus for preload display of object information are provided. The method includes following steps: capturing an external object information on a side of the transparent display opposite to a user by using the first information capturing device; pre-identifying at least one object in the external object information and tracing the at least one object according to an identification result; capturing information of the user by using the second information capturing device for pre-identifying and tracing a pointing range of the user, and determining whether the traced objects appear within the pointing range; and preload displaying object information of the objects appearing within the pointing range on the transparent display.

Abstract: This disclosure is enables various technologies involving various actions based on an object detecting a defined area and the defined area detecting the object.

Abstract: The present disclosure provides an image generating device configured to acquire an image captured by an imaging device to be installed in a water-surface movable body, generate a synthesized image where a graphic rendering a three-dimensional virtual reality object indicative of additional display information is synthesized with the captured image based on positional information, posture information, and the additional display information, estimate a position of the target object at a timing later than a timing at which the additional display information is acquired the latest, based on a speed of the target object included in the additional display information, or a speed of the target object estimated from the additional display information, and update the rendering of the virtual reality object of the target object in the synthesized image based on a result of the estimation.

Abstract: Embodiments of the present disclosure provide an image recognition method and apparatus, an electronic device, and a readable storage medium. The method for image recognition includes the steps of obtaining a to-be-recognized feature of a to-be-recognized image, the to-be-recognized image comprising a target person; obtaining a preset body model feature of a body frame image, the body model feature comprising locations of joints in the body frame image; inputting the to-be-recognized feature and the body model feature to a pre-constructed joint prediction model, the joint prediction model being obtained through training a neural network by using minimum respective differences between real locations of joints in a sample image and predicted locations of the corresponding joints as a training objective; and obtaining predicted locations of joints of the target person in the to-be-recognized image based on the joint prediction model.

Abstract: A processor may process a video feed of a monitored area. The processing may include attempting to decrypt the video feed using a temporally-varying digital rights management key in a state corresponding to a time at which the encrypted video feed was received and encountering a decryption error during the attempting. The processing may include comparing the video feed with an output feed from a secondary sensor to determine that at least one object is indicated in the monitored area by the output feed and not visible in the video feed. The processor may indicate a problem with the video feed. For example, the processor may indicate that the encrypted video feed has been altered prior to the receiving of the encrypted video feed due to the decryption error and/or that the video camera is malfunctioning due to the at least one object being not visible in the video feed.

Abstract: An approach is provided in which an information handling system trains a classifier using rated content segments that each has a first content type rating corresponding to a content type. Then, the information handling system uses the trained classifier to classify unrated content segments corresponding to an unrated content and generates second content type ratings for each of unrated content segments accordingly that identify a corresponding content type. In turn, the information handling system generates an overall content rating of the unrated content based on a combination of the second content type ratings.

Abstract: Disclosed herein are a method for evaluating social intelligence and an apparatus for the same. The method includes creating multiple segmented video clips by segmenting, based on behavior recognition, an observation video sequence that captures the social interaction behavior of the target to be evaluated; and evaluating the social intelligence of the target by calculating an evaluation score based on the similarities between ground truth, created based on social interaction analysis, and the multiple segmented video clips.

Abstract: Examples of a character recognition system are provided. In an example, the system may receive an object detection requirement pertaining to a video clip. The system may identify a visual media feature map from visual media data to process the object detection requirement. The system may implement an artificial intelligence component to segment the visual media feature map into a plurality of regions, and identify a plurality of image proposals therein. The system may implement a first cognitive learning operation to allocate a human face identity for a human face and an object name for an object present in the video clip. The system may determine a face identity model for the human face present in the plurality of image proposals and generate a tagged face identity model. The system may implement a second cognitive learning operation to assemble the plurality of frames with an appurtenant tagged face identity model.

Abstract: An electronic apparatus is provided. The electronic apparatus includes a communicator; and a processor configured to: receive multimedia data from an external apparatus through the communicator; identify a plurality of objects in a first frame among a plurality of frames included in the multimedia data; obtain content information corresponding to the plurality of objects in the first frame; obtain a first fingerprint based on the first frame; and control the communicator to transmit the content information and the first fingerprint to a server.

Abstract: A system and method for searching a video stream collected by a camera in a video surveillance system for an object's placement or displacement is disclosed. The searching includes an interactive question/answer approach that allows for a video snippet including the object's placement or displacement to be found quickly without the need for complicated video analytics. During the search, frames from algorithmically selected points in the video stream are presented to a user for review. The user reviews each frame and indicates if he/she sees the object. Based on the user's response the searching algorithmically reduces the portion video stream that is searched until a snippet of a video is found that includes the object's placement or displacement.

Abstract: A predictive security camera system includes a camera, an object recognizer for detecting an object in a current frame of the camera, a convolutional encoder for generating a feature representation of the current frame, a primary clustering module for clustering the feature representation in primary clusters of the detected object, a vector generation module for generating an information vector that includes a set of variables such as primary class, sub-class, and one or more external factors associated with the feature representation, an anomaly clustering module for clustering each information vector into one or more secondary clusters wherein each variable in the information vector is a clustering dimension, and a predictive analysis module for receiving at least one variable of an information vector as an input, and predicting at least one another variable of the information vector based on one or more previous secondary clusters formed by the anomaly clustering module.

Abstract: Systems and methods for recognizing fine-grained objects are provided. The system divides unlabeled training data from a target domain into two or more target subdomains using an attribute annotation. The system ranks the target subdomains based on a similarity to the source domain. The system applies multiple domain discriminators between each of the target subdomains and a mixture of the source domain and preceding target domains. The system recognizes, using the multiple domain discriminators for the target domain, fine-grained objects.

Abstract: A low-cost, low-power, stand-alone sensor platform having a visible-range camera sensor, a thermopile array, a microphone, a motion sensor, and a microprocessor that is configured to perform occupancy detection and counting while preserving the privacy of occupants. The platform is programmed to extract shape/texture from images in spatial domain; motion from video in time domain; and audio features in frequency domain. Embedded binarized neural networks are used for efficient object of interest detection. The platform is also programmed with advanced fusion algorithms for multiple sensor modalities addressing dependent sensor observations. The platform may be deployed for (i) residential use in detecting occupants for autonomously controlling building systems, such as HVAC and lighting systems, to provide energy savings, (ii) security and surveillance, such as to detect loitering and surveil places of interest, (iii) analyzing customer behavior and flows, (iv) identifying high performing stores by retailers.

Abstract: A vehicle ranging system includes a ranging scanner, a ranging data processor, and an object classifier. The ranging scanner may be configured to generate a series of ranging data. The ranging data processor may be configured to apply a scale invariant ranging template to the series of ranging data and determine a difference between the ranging template and the series of ranging data. The object classifier may be configured to identify an object corresponding to the series of ranging data based on the difference between the ranging template and the series of ranging data.

Abstract: According to one aspect, scene classification may be provided. An image capture device may capture a series of image frames of an environment from a moving vehicle. A temporal classifier may classify image frames with temporal predictions and generate a series of image frames associated with respective temporal predictions based on a scene classification model. The temporal classifier may perform classification of image frames based on a convolutional neural network (CNN), a long short-term memory (LSTM) network, and a fully connected layer. The scene classifier may classify image frames based on a CNN, global average pooling, and a fully connected layer and generate an associated scene prediction based on the scene classification model and respective temporal predictions. A controller of a vehicle may activate or deactivate vehicle sensors or vehicle systems of the vehicle based on the scene prediction.

Abstract: The present disclosure relates to a method for identifying an obstacle on a driving ground and a robot for implementing the same, and according to one embodiment of the present disclosure, the method for identifying an obstacle on a driving ground comprises the steps of allowing: a sensing module of the robot to sense the depths of objects in a driving direction so as to generate first depth information; a plane analysis unit of the sensing module to calibrate second depth information by using ground information stored by a map storage of the robot so as to generate the first depth information; a control unit to identify an obstacle from the second depth information; and the control unit to store position information of the identified obstacle in a temporary map of the map storage.

Abstract: Embodiments of the present disclosure provide a sensor calibration method, a sensor calibration device, a computer device, a storage medium, and a vehicle. The method includes: detecting surrounding objects in a travelling process of a vehicle; recognizing a static object from the surrounding objects; performing feature extraction on the static object by a camera and a lidar, respectively; and calibrating an extrinsic parameter of the camera and the lidar based on the extracted feature.

Abstract: An automobile head up display system and an obstacle prompt method thereof are provided. The method includes: acquiring a road condition video image, identifying obstacles from the acquired image after performing an enhancement processing on the acquired image; projecting and displaying a prompt information at corresponding positions of a front windshield of the automobile according to positions of the identified obstacles in the road condition video image.

Abstract: Systems, methods, and non-transitory computer-readable media can determine contextual information describing at least one physical structure corresponding to a location based at least in part on data captured by one or more sensors of a vehicle. A set of candidate interaction points for the at least one physical structure can be determined based at least in part on the determined contextual information describing the at least one physical structure corresponding to the location. The set of candidate interaction points can be filtered to identify one or more interaction points. An interaction point can be selected from the one or more interaction points to use for stopping the vehicle.

Abstract: Parking spot detection methods and systems that detect and pair parking line markers, improving the accuracy and speed of parking spot detection in vehicular driver assist (DA) and autonomous driving (AD) applications. These methods and systems incorporate three major steps: (1) preprocessing—one or more standard camera images are input into the algorithm and a bird's-eye-view (BEV) image is output from the algorithm; (2) deep neural network (DNN) segmentation of the parking line markers—the BEV image is input into the algorithm and a binary image is output from the algorithm (with, e.g., the parking line markers areas displayed in white and the background displayed in black); and (3) binary image processing—the binary image is input into the algorithm and detected and paired parking line markers representing the parking spots are output from the algorithm.

Abstract: A method and device for analyzing an image and providing the analysis for a driving assistance system of a vehicle, including recording the image; determining an operating state of the vehicle; analyzing the image using at least one image analysis method that is selected from at least two possible image analysis methods as a function of the operating state of the vehicle; and providing the analysis of the image as data values for the driving assistance system.

Abstract: A method and device of multi-focal sensing of an obstacle. A method includes acquiring detection results of obstacles at multiple moments by utilizing a camera with long focus lens and a camera with short focus lens; performing a target tracking to the acquired detection results, to obtain at least two tracking sequences, wherein each tracking sequence includes detection results acquired at the multiple moments for a same obstacle; and matching two random tracking sequences of the at least two tracking sequences, and combining the two random tracking sequences into a combined tracking sequence, if the matching is successful.

Abstract: Driver behavior recognition or driver behavior prediction are described herein. A first image sequence including image frames associated with a forward-facing image capture device of a vehicle and a corresponding vehicle data signal sequence may be received. A second image sequence including image frames associated with a rear or driver facing image capture device of the vehicle may be received. Feature vectors may be generated for respective sequences using neural networks, such as a convolutional neural network (CNN), a depth CNN, a recurrent neural network (RNN), a fully connected layer, a long short term memory (LSTM) layer, etc. A fusion feature may be generated by performing data fusion on any combination of the feature vectors. A predicted driver behavior may be generated based on the LSTM layer and n image frames on an image sequence and include x number of prediction frames.

Abstract: Methods and systems for biometric identification. Identification accuracy confidence is improved in one-to-many biometric matching by determining the mobile devices present in the area of the subject-to-be-identified, and determining the set of people associated with the mobile devices. Biometric profiles may be filtered using the set of people determined to be in the location of the subject-to-be-identified based on mobile device presence. The resulting subset provides higher confidence identification if a match is found with a profile belonging to a member of the subset.

Abstract: Anti-spoofing technology is provided for verifying a user of a fixed computer terminal. Image data of at least one verification image is received, as captured by an image capture device of the fixed computer terminal at a time corresponding to a request for access to a restricted function of the fixed computer terminal. User verification is applied to determine whether to grant access to the restricted function of the fixed computer terminal. A differential feature descriptor is determined, which encodes feature differences between the verification image data and image data of at least one unobstructed background image as captured by the image capture device. An anti-spoofing classifier processes the differential feature descriptor to classify it in relation to real and spoofing classes. Access to the restricted function of the fixed computer terminal is refused or granted based on the classification of the differential feature descriptor by the anti-spoofing classifier.

Abstract: Systems and methods are provided for performing automated face recognition and comparison. An input image of a face may be received and cropped, and the image may be processed through a deep neural network (DNN) to produce a k-dimensional feature vector. The k-dimensional feature vector may be converted to a k-dimensional binary vector by transforming each value in the vector to either 1 or 0. To search for nearest matches of the image in a database of gallery images of faces, the system may compare sub-strings of the binary vector to hash tables created from sub-strings of the gallery images, enabling sub-linear searching that allows locating the closest matches from among the entire gallery without requiring an exhaustive linear search of the entire gallery.

Abstract: A simulator to simulate target detection and recognition is provided. The simulator includes a memory configured to store instructions and a processor disposed in communication with said memory. The processor upon execution of the instructions is configured to receive ground truth attributes defining one or more respective simulated physical objects and receive collection attributes defining one or more collections. Each collection is a simulation of automated target recognition applied to simulated sensor output associated with simulated sensing of the one or more simulated physical objects. The processor upon execution of the instructions is further configured to simulate detection of the one or more simulated physical objects and recognition based on the ground truth attributes and the collection attributes and output detection data about at least one detected object based on the simulation of target detection.

Abstract: A method to simulate target detection and recognition is provided. The method includes receiving ground truth attributes defining one or more respective simulated physical objects and receiving collection attributes defining one or more collections. Each collection is a simulation of automated target recognition applied to simulated sensor output associated with simulated sensing of the one or more simulated physical objects. The method further includes simulating detection of the one or more simulated physical objects and recognition based on the ground truth attributes and the collection attributes, generating detection data about at least one detected object based on the simulated target detection, and outputting the detection data.

Abstract: Methods and systems for addressing malfunction of a medical imaging device are disclosed. The method includes classifying a type of an image artifact in a medical image acquired by the medical imaging device by using a trained machine learning model. The method also includes analyzing system data associated with acquisition of the medical image to identify one or more system parameters that might have contributed to the type of image artifact and providing an action for addressing the image artifact based on the identified one or more system parameters.

Abstract: A device for direct optical recording of skin prints offers recording of human skin prints for personal identification permitting a display layer directly below the placement surface. A light guide layer arranged below the sensor layer has at least one LED at a narrow side and light out-coupling structures which, by means of an inclination angle c and differences in the refractive indices relative to neighboring layers, permit a directed coupling out of light at a defined angle which results in total internal reflection at the placement surface at the air interface and with a small divergence angle range of ?+/?15°. A first adhesion layer between cover layer and sensor layer and a second adhesion layer between sensor layer and light guide layer are provided, the refractive indices of which are 1% to 30% lower than the refractive indices of light guide layer and sensor layer.

Abstract: Provided are character image processing methods and apparatuses, devices, storage medium, and computer programs. The character image processing method mainly comprises: obtaining at least one image block containing a character in a character image to be processed; obtaining image block form transformation information of the image block on the basis of a neural network, the image block form transformation information being used for changing a character orientation in the image block to a predetermined orientation, and the neural network being obtained by means of training using an image block sample having form transformation label information; performing form transformation processing on the character image to be processed according to the image block form transformation information; and performing character recognition on the character image to be processed which is subjected to the form transformation.

Abstract: The present disclosure provides automated robotic microscopy systems that facilitate high throughput and high content analysis of biological samples, such as living cells and/or tissues. In certain aspects, the systems are configured to reduce user intervention relative to existing technologies, and allow for precise return to and re-imaging of the same field (e.g., the same cell) that has been previously imaged. This capability enables experiments and testing of hypotheses that deal with causality over time with greater precision and throughput than conventional microscopy methods.

Abstract: A data capture component receives a video stream comprising a plurality of frames, wherein each frame comprises a data field. One or more text regions in a selected frame of the plurality of frames are identified. One of the one or more identified text regions that corresponds to a set of attributes associated with the data field are selected. The data of the one of the one or more identified text regions of the selected frame are compared with data of one or more text regions of a subsequent frame. Responsive to determining that the data of the one or more text regions of the subsequent frame is a closer match to the set of attributes, the data of the one of the one or more identified text regions of the selected frame are updated. The data of the one of the one or more identified text regions is then provided to a client device.

Abstract: An image processing apparatus, an image processing method, and a program, provide accurate collation even when an image contains a number of identical or similar subjects. The image processing apparatus generates, with respect to feature points to be detected from a first image, a first local feature amount group including local feature amounts representing feature amounts of local regions containing the respective feature points, and a first coordinate position information group including coordinate position information. The image processing apparatus clusters the feature points of the first image based on the first coordinate position information group. The image processing apparatus collates, in units of clusters, the first local feature amount group with a second local feature amount group formed from local feature amounts of feature points detected from a second image.

Abstract: Information about a drug-containing vessel is determined by capturing image data of the curved surface of a cylindrical portion of a drug-containing vessel. The image data is unfurled from around the curved surface, binarised, and a template matching algorithm employed to determine that the label information comprises candidate information about the vessel and/or the drug.

Abstract: A method for vehicle damage identification, includes: obtaining a vehicle damage picture to be identified; inputting the vehicle damage picture into a plurality of pre-trained target detection models respectively, and obtaining corresponding detection results from the plurality of target detection models as a detection result set, wherein the detection result set comprises candidate bounding boxes detected by the plurality of target detection models and category prediction results of the candidate bounding boxes; determining an integrated feature vector of a first candidate bounding box of the candidate bounding boxes; and separately inputting integrated feature vectors corresponding to the candidate bounding boxes into a pre-trained classification model, and optimizing the detection result set according to output results of the classification model.

Abstract: A system for displaying one or more images to select one or more matching formulas to match color and appearance of a target coating of an article includes a computing device, a display device, a host computer connected to the computing device, one or more data input devices, a first database containing repair formulas, color characteristics, and appearance characteristics, a second database containing identification information of an article or a three-dimensional model of an article or a three-dimensional mapping of a geometry of a part of an article, and a computer program product accessible to the computing device and/or the host computer and performing a computing process to retrieve at least one preliminary matching formula from the first database, select one article or a three-dimensional mapping of a geometry of one article from the second database, generate individual matching images, and display the individual matching images on the display device.

Abstract: A device configured to link a first string correlithm object to a second correlithm object in a correlithm object processing system, includes a node configured to receive a particular sub-string correlithm object in the first string correlithm object. The node determines a first proximity in n-dimensional space from the particular sub-string correlithm object to a first sub-string correlithm object in the second string correlithm object. The node determines a second proximity in n-dimensional space from the particular sub-string correlithm object to a second sub-string correlithm object in the second string correlithm object. The node maps the particular sub-string correlithm object to a selected one of the first or second sub-string correlithm object of the second correlithm object based upon the closest of the first proximity and the second proximity.

Abstract: A method for detecting temporal pattern anomalies in a video stream includes detecting an object in a current frame of the video stream, generating a processed current frame that contains the detected object, generating a feature representation of the processed current frame, clustering the feature representation in one or more primary clusters in a clustering space of the primary class, generating an information vector of the feature representation, that includes information regarding the primary class, the sub-class and one or more external factors associated with the feature representation, clustering each information vector into one or more secondary clusters, and reporting a next frame as an anomaly when a corresponding information vector is positioned outside a secondary cluster of a feature presentation of a previous frame.

Abstract: An information processing system includes: a recognition unit configured to perform recognition processing on a video captured by an image capturing unit; a determination unit configured to determine timing to update a model used for the recognition processing; and an update unit configured to update the model at the determined timing.

Abstract: Disruption of computerized face detection includes receiving a source image that contains a representation of a face and computing a perturbation for the source image. The perturbation is specific to the source image and is configured for a target face detector. A perturbed image is then generated by adding the perturbation to the source image and then the perturbed image may be outputted instead of the source image.

Abstract: Concepts and technologies disclosed herein are directed to data harvesting for machine learning model training. According to one aspect of the concepts and technologies disclosed herein, a network data collection system can identify a target data source location from which to harvest data for a machine learning system to utilize during a machine learning model training process. The data can be associated with a plurality of mobile communications devices operating in communication with at least one base station of a mobile communications network that serves the target data source location. The network data collection system can collect the data and provide the data to the machine learning system. The machine learning system, in turn, can create a training data set for use during the machine learning model training process based, at least in part, upon the data.

Abstract: A method for generating training examples for a product recognition model is disclosed. The method includes capturing images of a product using an array of cameras. A product identifier for the product is associated with each of the images. A bounding box for the product is identified in each of the images. The bounding boxes are smoothed temporally. A segmentation mask for the product is identified in each bounding box. The segmentation masks are optimized to generate an optimized set of segmentation masks. A machine learning model is trained using the optimized set of segmentation masks to recognize an outline of the product. The machine learning model is run to generate a set of further-optimized segmentation masks. The bounding box and further-optimized segmentation masks from each image are stored in a master training set with its product identifier as a training example to be used to train a product recognition model.

Abstract: Systems and methods for training image processing neural networks by synthetic photorealistic indicia-bearing images. An example method comprises: generating an initial set of images, wherein each image of the initial set of images comprises a rendering of a text string; producing an augmented set of images by processing the initial set of images to introduce, into each image of the initial set of image, at least one simulated image defect; generating a training dataset comprising a plurality of pairs of images, wherein each pair of images comprises a first image selected from the initial set of images and a second image selected from the augmented set of images; and training, using the training dataset, a convolutional neural network for image processing.

Abstract: Various techniques are described for automatically suggesting variation parameters used to generate a tailored synthetic dataset to train a particular machine learning model. A seeding taxonomy associates a plurality of machine learning scenarios with corresponding subsets of variation parameters. A selected machine learning scenario is used to retrieve a corresponding subset of variation parameters associated with the selected machine learning scenario by the seeding taxonomy. The seeding taxonomy may be adaptable using a feedback loop that tracks selected variation parameters and updates the seeding taxonomy. The suggested variation parameters are presented as suggestions to assist users to identify and select relevant variation parameters faster and more efficiently. Further embodiments relate to pre-packaging synthetic datasets for common or anticipated machine learning scenarios.

Abstract: Various techniques are described for monitoring progress of a distributed Synthetic Data as a Service (SDaaS) in processing a synthetic data request. A progress portal can be provided that allows a requesting account to actively monitor SDaaS progress in processing a request and to issue commands to assign a resource to process an associated synthetic data task, even after beginning to process the request. The portal may support commands to assign or change resources (e.g., CPUs/GPUs), batch the processing of a synthetic data task, maximize or minimize a derived processing characteristic (e.g., processing time, cost, quality of output, etc.), set the size or quality of generated synthetic data assets, and the like. The portal may provide various types of feedback, such as progress in processing the request, CPU/GPU usage, a measure of time spent on synthetic data creation by a particular type of processor, and the like.