A method and system of comparing at least two digital labels is disclosed. A processor detects a set of regions of interest (ROIs) corresponding to a set of objects in each of the at least two digital labels using a first Machine Learning (ML)/Deep Learning (DL) model. Each of the set of objects are classified into one of a set of text objects or a set of non-text objects. One or more key-value pairs of one or more attributes are extracted from the text object. The one or more key-value pairs text data are compared corresponding to the text object of the at least two digital labels. The each of non-text object of the set of non-text objects are categorized into an object category from a set of object categories. An ROI of the each of non-text object of the set of non-text objects may be compared.

One aspect of the present disclosure provides a method for estimating attributes of a person in an image, the method comprising detecting an object region including a whole body region, a visible body region, and a head region of at least one person in an input image, determining whether to estimate attributes of the person based on at least one of a relative position of the head region with respect to the whole body region, or a ratio of an overlapping region between the whole body region and the visible body region, and estimating the attributes of the person based on the input image, when it is determined to estimate the attributes of the person.

Provided is an external environment recognition device capable of determining a place where a host vehicle can safely evacuate at the time of recognition of a specific vehicle by using a plurality of cameras in combination even in a vehicle not equipped with a distance sensor such as a radar, a LiDAR, an ultrasonic sensor, or an infrared sensor. Therefore, an external environment recognition device includes: a plurality of cameras installed so as to have a plurality of stereo vision regions in which at least a part of a visual field region overlaps around a host vehicle; a three-dimensional information generation unit that generates three-dimensional information by performing stereo matching processing in each of the plurality of stereo vision regions; a three-dimensional information accumulation unit that accumulates the three-dimensional information generated during traveling of the host vehicle in time series; and a three-dimensional information update unit that updates the three-dimensional information accumulated in the three-dimensional information accumulation unit using three-dimensional information newly generated by the three-dimensional information generation unit.

Disclosed is an automatic color matching configuration method and device for marble, where the method includes the following steps: obtaining a to-be-repaired image of to-be-repaired marble for image recognition to determine a to-be-repaired contour in the to-be-repaired image; extending outward a preset distance along the to-be-repaired contour to generate a reference contour, and determining an area between the reference contour and the to-be-repaired contour as a reference area; filling a base coating into a to-be-repaired area in the to-be-repaired contour to form an initial base coat; determining a base pixel value based on a color ratio of the reference area, and performing base color matching on the initial base coat based on the base pixel value to obtain a current base coat.

A camera with a field of view toward an external environment of an aircraft is disposed within an aircraft door such that a jet bridge cabin is within the field of view of the camera when the aircraft is within a docking distance of the jet bridge cabin. A display device is disposed within an interior of the aircraft. A processor is operatively coupled to the camera and to the display device. The processor analyzes image data captured by the camera by identifying physical characteristics of the jet bridge cabin, extracting, using the identified physical characteristics of the jet bridge cabin, alignment features corresponding to the physical characteristics of the jet bridge cabin that are indicative of alignment between the jet bridge cabin and the aircraft door, determining, based on the alignment features, whether the physical characteristics of the jet bridge cabin within the captured image data satisfy threshold alignment criteria to produce an alignment state, and outputting an indication of the alignment state.

Various methods and systems are provided for identifying future occurrences of lesions in a patient. In one example, a method includes feeding first images collected from one or more patients prior to appearance of lesions and second images collected from the one or more patients after appearance of the lesions to a processor to train a model to predict a location of a future lesion. The method further includes inputting third images collected from a new patient to the processor to infer regions for future lesions, and displaying the inferred regions in a probability map at a display unit to indicate areas of increased likelihood of lesion occurrence.

An image processing apparatus includes an extractor and an object identifier. The extractor extracts a left feature quantity included in a left image of a stereo image and a right feature quantity included in a right image of the stereo image by executing a convolution calculation multiple times on the basis of the left and right images with securing horizontal symmetry between the left feature quantity and the right feature quantity, and executes, at the convolution calculation, a concatenation process of concatenating pixel values between the left feature quantity and the right feature quantity using a calculation process to secure the horizontal symmetry. The object identifier performs object identification of an object on the basis of the left feature quantity and the right feature quantity.

A method may include generating views materializing tensors generated by a convolutional neural network operating on an image. Determining the outputs of the convolutional neural network operating on the image with a patch occluding various portions of the image. The outputs being determined by generating queries on the views that performs, based at least on the changes associated with occluding different portions of the image, partial re-computations of the views. A heatmap may be generated based on the outputs of the convolutional neural network. The heatmap may indicate the quantities to which the different portions of the image contribute to the output of the convolutional neural network operating on the image. Related systems and articles of manufacture, including computer program products, are also provided.

A method of training an object detector including obtaining a first tracklet set based on an object detection result output corresponding to a plurality of frames, obtaining a second tracklet set from ground truth data predetermined corresponding to the plurality of frames, obtaining a first bipartite matching result of a bounding box level, the bounding box level corresponding to each of first tracklets included in the first tracklet set and each of second tracklets included in the second tracklet set, obtaining a second bipartite matching result of a tracklet level, the tracklet level corresponding to the first tracklet set and the second tracklet set, based on the first bipartite matching result, and assigning a second tracklet determined to be one of a pair including a first tracklet, as a paired first tracklet and second tracklet, to ground truth data of the first tracklet, based on the second bipartite matching result.

A data processing method includes inputting first blocks from a plurality of blocks of a divided to-be-processed image into an image processing model, to obtain block processing results of the first blocks, and performing splicing processing on the block processing results of the first blocks and block processing results of second blocks of the processed image, to obtain an image processing result of the to-be-processed image. The first blocks are blocks of the to-be-processed image, which satisfy a processing condition. The processed image and the to-be-processed image are divided in a same division method. The second blocks are blocks of the to-be-processed image which do not satisfy the processing condition, compared to the processed image.