The present disclosure relates to an image processing method and apparatus, an electronic device, and a storage medium. The method includes: acquiring at least two target images; determining an attention map of at least one target in each of the at least two target images according to a result of detecting target of each target image, where the attention map indicates the position of one target in a target image; and determining, based on each target image and the attention map of the at least one target in the each target image, a result of association that indicates whether a correspondence exists between at least some of targets in different target images.

A system and method for predicting driving actions based on intent-aware driving models that include receiving at least one image of a driving scene of an ego vehicle. The system and method also include analyzing the at least one image to detect and track dynamic objects located within the driving scene and to detect and identify driving scene characteristics associated with the driving scene and processing an ego-thing graph associated with the dynamic objects and an ego-stuff graph associated with the driving scene characteristics. The system and method further include predicting a driver stimulus action based on a fusion of representations of the ego-thing graph and the ego-stuff graph and a driver intention action based on an intention representation associated with driving intentions of a driver of the ego vehicle.

Disclosed is a ship identity recognition method based on the fusion of AIS data and video data, comprising: collecting a ship sample to train a ship target classifier; performing, using the ship target classifier, ship target detection on a video frame collected by a gimbal camera; performing a comparison with a recognized ship library to filter a recognized ship; acquiring AIS data and filtering same across time and spatial scales; predicting the current position of an AIS target using a linear extrapolation method and converting the current position to an image coordinate system; performing position matching between a target to be matched and the converted AIS target; and performing feature extraction on the successfully matched target and storing the extracted feature, together with ship identity information, into the recognized ship library. Experimental results show that the present invention can quickly and accurately extract a surveillance video and perform identity recognition on the ship target, effectively reduces labor costs, and has a broad application prospect in the fields such as ship transportation and port management.

A learning apparatus includes circuitry. The circuitry receives an input of first label information to be given to a facial expression image indicating a face of a person. The circuitry estimates second label information to be given to the facial expression image based on an interpolated image generated using the facial expression image and line-of-sight information indicating a direction of a line of sight of an annotator, the direction being detected at a time when the input is received. The circuitry calculates a difference between the first label information of which the input is received and the estimated second label information. The circuitry updates a parameter used for processing of estimating the second label information based on the calculated difference.

The invention discloses a method for glass detection in a real scene, which belongs to the field of object detection. The present invention designs a combination method based on LCFI blocks to effectively integrate context features of different scales. Finally, multiple LCFI combination blocks are embedded into the glass detection network GDNet to obtain large-scale context features of different levels, thereby realize reliable and accurate glass detection in various scenarios. The glass detection network GDNet in the present invention can effectively predict the true area of glass in different scenes through this method of fusing context features of different scales, successfully detect glass with different sizes, and effectively handle with glass in different scenes. GDNet has strong adaptability to the various glass area sizes of the images in the glass detection dataset, and has the highest accuracy in the field of the same type of object detection.

The subject disclosure presents systems and computer-implemented methods for assessing a risk of cancer recurrence in a patient based on a holistic integration of large amounts of prognostic information for said patient into a single comparative prognostic dataset. A risk classification system may be trained using the large amounts of information from a cohort of training slides from several patients, along with survival data for said patients. For example, a machine-learning-based binary classifier in the risk classification system may be trained using a set of granular image features computed from a plurality of slides corresponding to several cancer patients whose survival information is known and input into the system. The trained classifier may be used to classify image features from one or more test patients into a low-risk or high-risk group.

A synthesizing apparatus comprises: an input part that inputs a plurality of feature point sets that are respectively extracted by a plurality of methods from an input image having a curved stripes pattern formed by ridges; and a synthesizing part that synthesizes the plurality of feature point sets by executing a logical operation on the plurality of feature point sets. The synthesizing part can execute a logical OR operation on the plurality of feature point sets. The synthesizing part can also execute a logical AND operation on the plurality of feature point sets.

An object detection apparatus comprises circuitry configured to obtain one or more feature maps of a measurement space generated from sensor data of a scene, the sensor data including position information and a feature map representing feature values of a feature at a plurality of positions in the measurement space, transform the one or more feature maps into a template coordinate system, compute a likelihood that the one or more feature maps correspond to an object template and a given candidate object configuration, and iteratively compute likelihoods for different candidate object configurations and determine the object configuration with the highest likelihood.

Example aspects of the present disclosure are directed to computing systems and methods that perform whole person association with face screening and/or face hallucination. In particular, one aspect of the closure is directed to a multi-headed person and face detection model that performs both face and person detection in one model. Each of the face and person detection can find landmarks or other pose information and also a confidence score. The pose information for the face and person detections can be used to select certain face and person detections to associate together as a whole person detection, which can be referred to as an “appearance.”

A method for assessing and inspection of wind turbine blades 4, in particular moving wind turbine blades, comprising the steps of directing a data capture device such as a camera system 1 towards a wind turbine blade 4 that is to be assessed. The camera system 1 can be attached to an aerial craft such as a helicopter 3, and is provided with a laser 13 that is used to track the motion of the blade 4 that is to be assessed. The laser 13 may be adapted to track a single blade 4 or the camera system 1 may be provided with multiple lasers to track multiple blades of the same turbine at the same time. The method further comprises collecting data of the state or condition of the blade 4 using the camera system 1 during the time that the helicopter 3 navigates around the wind turbine 2. The image data of the blade that is captured is fed into a computer processor (not shown) which can be on-board the helicopter 3 or at a remote location. The computer processor is adapted to reconstruct the image data into a 2-D or 3-D virtual digital image of the wind turbine 2. The method further comprises using at least one algorithm to compare and contrast various parts of the digital image generated by the reconstruction, with corresponding parts of a predetermined image of a healthy wind turbine, to identify defects or damage to the actual wind turbine, and the extent of the defects and damage. Using machine learning and A.I., the method is able to ascertain if and when replacement of the wind turbine blade may be necessary. An apparatus for undertaking the method is also claimed.