An apparatus, method, system and computer readable medium for video tracking. An exemplar crop is selected to be tracked in an initial frame of a video. Bayesian optimization is applied with each subsequent frame of the video by building a surrogate model of an objective function using Gaussian Process Regression (GPR) based on similarity scores of candidate crops collected from a search space in a current frame of the video. A next candidate crop in the search space is determined using an acquisition function. The next candidate crop is compared to the exemplar crop using a Siamese neural network. Comparisons of new candidate crops to the exemplar crop are made using the Siamese neural network until the exemplar crop has been found in the current frame. The new candidate crops are selected based on an updated surrogate model.

Forensic method for identifying forged documents. For each of a stream of incoming jpeg images, using a processor configured for determining whether jpeg image/s is a replacement forgery by determining whether a first portion of individual image which resides at a known location (known likely to be replaced by forger) within the individual jpeg image has been replaced, including: indicator, face-djpg, for the first portion at known location; computing indicator, aka nonface-djpg, for a second portion of individual image which resides at a comparison location within the jpeg image known as unlikely to be replaced by a forger; and determining whether face-djpg and nonface-djpg fulfill predetermined logical criterion and deciding whether the individual jpeg image is a replacement forgery accordingly.

The present disclosure provides a saliency prediction method and system for a 360-degree image based on a graph convolutional neural network. The method includes: firstly, constructing a spherical graph signal of an image of an equidistant rectangular projection format by using a geodesic icosahedron composition method; then inputting the spherical graph signal into the proposed graph convolutional neural network for feature extraction and generation of a spherical saliency graph signal; and then reconstructing the spherical saliency graph signal into a saliency map of an equidistant rectangular projection format by using a proposed spherical crown based interpolation algorithm. The present disclosure further proposes a KL divergence loss function with sparse consistency. The method can achieve excellent saliency prediction performance subjectively and objectively, and is superior to an existing method in computational complexity.

The present disclosure provides a saliency prediction method and system for a 360-degree image based on a graph convolutional neural network. The method includes: firstly, constructing a spherical graph signal of an image of an equidistant rectangular projection format by using a geodesic icosahedron composition method; then inputting the spherical graph signal into the proposed graph convolutional neural network for feature extraction and generation of a spherical saliency graph signal; and then reconstructing the spherical saliency graph signal into a saliency map of an equidistant rectangular projection format by using a proposed spherical crown based interpolation algorithm. The present disclosure further proposes a KL divergence loss function with sparse consistency. The method can achieve excellent saliency prediction performance subjectively and objectively, and is superior to an existing method in computational complexity.

Provided are a method, apparatus and device for image feature extraction and network training. The method includes the following. A first association graph including a main node and at least one neighbor node is acquired. A node value of the main node represents an image feature of a target image. A node value of each of the at least one neighbor node represents an image feature of a respective one of at least one neighbor image. The at least one neighbor image is similar to the target image. The first association graph is input into a feature update network. The feature update network updates the node value of the main node according to the node value of the at least one neighbor node in the first association graph, to obtain an updated image feature of the target image.

Provided are a method, apparatus and device for image feature extraction and network training. The method includes the following. A first association graph including a main node and at least one neighbor node is acquired. A node value of the main node represents an image feature of a target image. A node value of each of the at least one neighbor node represents an image feature of a respective one of at least one neighbor image. The at least one neighbor image is similar to the target image. The first association graph is input into a feature update network. The feature update network updates the node value of the main node according to the node value of the at least one neighbor node in the first association graph, to obtain an updated image feature of the target image.

The present disclosure describes a computer-implemented method for image landmark detection. The method includes receiving an input image for the image landmark detection, generating a feature map for the input image via a convolutional neural network, initializing an initial graph based on the generated feature map, the initial graph representing initial landmarks of the input image, performing a global graph convolution of the initial graph to generate a global graph, where landmarks in the global graph move closer to target locations associated with the input image, and iteratively performing a local graph convolution of the global graph to generate a series of local graphs, where landmarks in the series of local graphs iteratively move further towards the target locations associated with the input image.

An image processing technique is presented using a hierarchical image model. The technique may be used as a precursor to subsequent image processing, to fix detected images in a post processing stage or as a segmentation or classification stage. The techniques may also be applied to super resolution. In a first layer of the hierarchical image model, each observed pixel of the image has a representation allocated to one or more input node. A set of the input nodes are assigned to a hidden node of a second layer, and a duplicate set of input nodes of the first layer is assigned to a duplicate of the hidden node in the second layer. In this way, a dense latent tree is formed in which a subtree is duplicated. Variables are assigned to the input nodes, the hidden node and the duplicate nodes and recurringly modified to process the image. Belief propagation messages may be utilised to recursively modify the variables. An image processing system using the method is described. A planning system for an autonomous vehicle comprising the image processing system is described.

Disclosed are a method for measuring the length of a living tissue included in a slide image, and a computing system for performing same. According to one aspect of the present invention, the method comprising the steps of: segmenting the slide image into a plurality of patches having a predetermined size; generating a graph corresponding to the slide image; for each edge included in the graph, setting a weight of the edge; for each connected component of the graph including two or more nodes, detecting shortest paths between all node pairs included in the connected components and determining a longest shortest path having the longest length from among the detected shortest paths between all the node pairs; and calculating the length of the living tissue included in the slide image, on the basis of the longest shortest path of each connected component constituting the graph.

Disclosed are a method for measuring the length of a living tissue included in a slide image, and a computing system for performing same. According to one aspect of the present invention, the method comprising the steps of: segmenting the slide image into a plurality of patches having a predetermined size; generating a graph corresponding to the slide image; for each edge included in the graph, setting a weight of the edge; for each connected component of the graph including two or more nodes, detecting shortest paths between all node pairs included in the connected components and determining a longest shortest path having the longest length from among the detected shortest paths between all the node pairs; and calculating the length of the living tissue included in the slide image, on the basis of the longest shortest path of each connected component constituting the graph.