A face verification method and apparatus is disclosed. The face verification method includes selecting a current verification mode, from among plural verification modes, to be implemented for the verifying of the face, determining one or more recognizers, from among plural recognizers, based on the selected current verification mode, extracting feature information from information of the face using at least one of the determined one or more recognizers, and indicating whether a verification is successful based on the extracted feature information.

In variants, a method for item recognition can include: optionally calibrating a sampling system, determining visual data using the sampling system, determining a point cloud, determining region masks based on the point cloud, generating a surface reconstruction for each item, generating image segments for each item based on the surface reconstruction, and determining a class identifier for each item using the respective image segments.

The disclosure provides an eye state assessment method and an electronic device. The method includes: obtaining an optic disc image area from a first fundus photography and generating multiple optic cup-to-disc ratio assessment results by multiple first models based on the optic disc image area; obtaining a first assessment result of an eye based on the optic cup-to-disc ratio assessment results; performing multiple data augmentation operations on the first fundus photography to generate multiple second fundus photographies; generating multiple retinal nerve fiber layer (RNFL) defect assessment results by multiple second models based on the second fundus photographies; obtaining a second assessment result of the eye based on the RNFL defect assessment results; and obtaining an optic nerve assessment result of the eye based on the first assessment result and the second assessment result.

The present disclosure relates to a method for recognizing facial expressions based on adversarial elimination. First, a facial expression recognition network is built based on a deep convolutional neural network. On a natural facial expression data set, the facial expression recognition network is trained through a loss function to make facial expression features easier to distinguish. Then some key features of input images are actively eliminated by using an improved confrontation elimination method to generate a new data set to train new networks with different weight distributions and feature extraction capabilities, forcing the network to perform expression classification discrimination based on more features, which reduces the influence of interference factors such as occlusion on the network recognition accuracy rate, and improving the robustness of the facial expression recognition network. Finally, the final expression classification predicted results are obtained by using network integration and a relative majority voting method.

A system such as an autonomous vehicle’s perception system will identify and classify an obstruction in a field of view of an image capturing device. The system will receive a sequence of image frames from the image capturing device. For each of the image frames, the system will segment the image frame into a regions of interest (ROIs), and the system will use a classifier to assign a classification to each ROI. The classification indicates whether the ROI is clear or obstructed. The system will aggregate the classifications for each ROI to determine an aggregate classification. When an obstructed classification persists for a threshold number of image frames, the system will classify the image capturing device as obstructed, and it will generate a function request that, when executed, will cause a system of which the image capturing device is a component to perform a function.

This disclosure relates to methods, non-transitory computer readable media, and systems that can classify term sequences within a source text based on textual features analyzed by both an implicit-class-recognition model and an explicit-class-recognition model. For example, by applying machine-learning models for both implicit and explicit class recognition, the disclosed systems can determine a class corresponding to a particular term sequence within a source text and identify the particular term sequence reflecting the class. The dual-model architecture can equip the disclosed systems to apply (i) the implicit-class-recognition model to recognize implicit references to a class in source texts and (ii) the explicit-class-recognition model to recognize explicit references to the same class in source texts.

A method for detecting a road-boundary based on radar information and visual information, the method may include (a) estimating, based on the visual information obtained by a visual sensor, locations of road-boundary points up to a first distance from the visual sensor; wherein a distance ambiguity of visual based location determination of road-boundary points with the first distance range does not exceed a predefined ambiguity threshold; wherein a distance ambiguity of visual based location determination of road-boundary points outside the first distance exceeds a predefined ambiguity threshold; (b) estimating, based at least on (i) the radar information, and (ii) angular-constrains regarding angular relationships between adjacent road-boundary points, locations of road-boundary points outside the first distance range; and (c) determining the shape and position of the road-boundary based on the locations of the road-boundary points within the first range distance.

A method for categorizing biological structure of interest (BSOI) in digitized images of biological tissues comprises a stage of identifying BSOIs in digitized images and further comprises presenting an image from the plurality of images that comprises at least one BSOI with high level of entropy to a user, receiving from the user input indicative of a category to be associated with the BSOI that had the high level of entropy and updating the cell categories classifier according to the category of the BSOI provided by the user.

A tiered processing scheme for processing image data is provided. A method can include obtaining image data indicative of one or more court judgment(s) with a number of features. The method can include obtaining judgment information from the image data by applying a number of image processing techniques in accordance with a processing hierarchy tailored to the image data. The image data can be classified by a machine-learning model and the processing hierarchy can be determined based on the classification. The processing hierarchy balances the computing resources used by a respective technique with the accuracy afforded by the technique when applied to image data with a respective classification. A computing system can utilize the processing hierarchy to leverage different image processing techniques in a tiered processing scheme tailored to image data.

An image object classification optimizing method and system are disclosed. The method is executed by a processor coupled to a memory. The method includes steps: providing an image file including at least one image object; performing a process of characteristics enhancement on the image object; performing a process of characteristics classification on the enhanced image object by an odd number of two-dimensional masks whose sizes are sequentially doubled, based on a plurality of characteristic parameters of a preferred classification model, to generate a plurality of classification results; and estimating variabilities of the plurality of classification results, sorting the variabilities, and selecting at least one of the classification results whose variability is lower than a variation tolerance as at least one optimization result, according to the sorting result.