A hologram detection method according to an aspect of the disclosure, includes: inputting a first image, obtained by capturing a detection object on the basis of first flash intensity, to a neural network model to obtain a first detection result value representing the detection or not of a hologram for each of predetermined at least one detection unit regions; and comparing a threshold value with the first detection result value obtained for each detection unit region to determine the detection or not of a hologram in the first image and a first detection unit region where a hologram is detected.

A system, method and computer-readable medium are provided to enable digital bank endorsement. A digital image of a back side of a check may be placed in a computer memory. Appropriate coordinates for a bank endorsement may be determined. A bank endorsement may be automatically generated. The digital image may then be electronically altered by overlaying, merging, or rendering text of the generated bank endorsement. A modified digital image may be combined with an image of the front side of the check and stored and/or exported to check clearing operations.

Implementations are described herein for analyzing a sequence of digital images captured by a mobile vision sensor (e.g., integral with a robot), in conjunction with information (e.g., ground truth) known about movement of the vision sensor, to determine spatial dimensions of object(s) and/or an area captured in a field of view of the mobile vision sensor. Techniques avoid the use of visual indicia of known dimensions and/or other conventional tools for determining spatial dimensions, such as checkerboards. Instead, techniques described herein allow spatial dimensions to be determined using less resources, and are more scalable than conventional techniques.

Approaches are described for shape-based graphics search. Each graphics object of a set of graphics objects is analyzed. The analyzing includes determining an outline of the graphics object from graphics data that forms the graphics object. The outline of the graphics object is sampled resulting in sampled points that capture the outline of the graphics object. A shape descriptor of the graphics object is determined which captures local and global geometric properties of the sampled points. Search results of a search query are determined based on a comparison between a shape descriptor of a user identified graphics object and the shape descriptor of at least one graphics object of the set of graphics objects. At least one of the search results can be presented on a user device associated with the search query.

A prediction system for simulating effects of a real-world event can be used for autonomous driving. In operation, the system receives input data regarding a complex system (e.g., roadways) and various real-world events. A full-scale network is constructed of the complex system, such that nodes represent road intersections and edges between nodes represent road segments linking the road intersections. The network is reduced is scaled down to generate a multi-layer model of the complex system. Each layer in the model is simulated to identify equilibrium flows, with the model thereafter destabilized by applying stimuli to reflect the real-world event. An autonomous vehicle can then be caused to chart and traverse a road path based on road segments and intersections that are least affected by the real-world event.

Systems and methods are provided for enhanced animation generation based on using motion mapping with local bone phases. An example method includes accessing first animation control information generated for a first frame of an electronic game including local bone phases representing phase information associated with contacts of a plurality of rigid bodies of an in-game character with an in-game environment. Executing a local motion matching process for each of the plurality of local bone phases and generating a second pose of the character model based on the plurality of matched local poses for a second frame of the electronic game.

A technique for improving the computational time for performing a non-maximum suppression operation may include receiving a request to perform a non-maximum suppression operation on a set of candidate predictions of a computing task, and performing a statistical analysis on a set of confidence scores corresponding to the set of candidate predictions to determine a standard deviation of the set of confidence scores. A confidence score threshold can be determined based on the standard deviation. Candidate predictions having a confidence score below the confidence score threshold can then be discarded to form a reduced set of candidate predictions. Additional candidate predictions can be discarded from the reduced set of candidate predictions based on an intersection-over-union overlap metric, and the remaining candidate predictions from the reduced set of candidate predictions can be provided as a result of the non-maximum suppression operation.

A technique for improving the computational time for performing a non-maximum suppression operation may include receiving a request to perform a non-maximum suppression operation on a set of candidate predictions of a computing task, and performing a statistical analysis on a set of confidence scores corresponding to the set of candidate predictions to determine a standard deviation of the set of confidence scores. A confidence score threshold can be determined based on the standard deviation. Candidate predictions having a confidence score below the confidence score threshold can then be discarded to form a reduced set of candidate predictions. Additional candidate predictions can be discarded from the reduced set of candidate predictions based on an intersection-over-union overlap metric, and the remaining candidate predictions from the reduced set of candidate predictions can be provided as a result of the non-maximum suppression operation.

This application disclose a method for training a model performed at a computing device. The method includes: acquiring a template image and a test image; invoking a first object recognition model to process a feature of a tracked object in the template image to obtain a first reference response, and a second object recognition model to process the feature in the template image to obtain a second reference response; invoking the first model to process a feature of a tracked object in the test image to obtain a first test response, and the second model to process the feature to obtain a second test response; tracking the first test response to obtain a tracking response of the tracked object; and updating the first object recognition model based on differences between the first and second reference responses, that between the first and second test responses, and that between a tracking label and the tracking response.

This application disclose a method for training a model performed at a computing device. The method includes: acquiring a template image and a test image; invoking a first object recognition model to process a feature of a tracked object in the template image to obtain a first reference response, and a second object recognition model to process the feature in the template image to obtain a second reference response; invoking the first model to process a feature of a tracked object in the test image to obtain a first test response, and the second model to process the feature to obtain a second test response; tracking the first test response to obtain a tracking response of the tracked object; and updating the first object recognition model based on differences between the first and second reference responses, that between the first and second test responses, and that between a tracking label and the tracking response.