US 2020/0026987 A1 · NEURAL NETWORK BASED POSITION ESTIMATION OF TARGET OBJECT OF INTEREST IN VIDEO FRAMES

Abstract

Visual target tracking is task of locating a target in consecutive frame of a video. Conventional systems observe target behavior frames of the video. However, dealing with this problem is very challenging when video has illumination variations, occlusion, change in size and view of the object due to relative motion between camera and object. Embodiments of the present disclosure addresses this problem by implementing Neural Network (NN), its features and their corresponding gradients. Present disclosure explicitly guides the NN by feeding target object of interest (ToI) defined by a bounding box in the first frame of the video. With this guidance, NN generates target activation map via convolutional features map and their gradient maps, thus giving tentative location of the ToI to further exploit to locate target object precisely by using correlation filter(s) and peak location estimator, thus repeating process for every frame of video to track ToI accurately.

Claims18

Dependent claims

A processor implemented method, comprising:

receiving, at a neural network, a first input comprising a video frame, and a second input comprising a target object of interest in the video frame, wherein a corresponding class label is generated based on the second input (502);
receiving, at a first part of the neural network, a current frame to generate one or more corresponding convolutional feature maps (504);
generating, by a softmax layer of the neural network, a probability score for the generated one or more corresponding convolutional feature maps (506);
generating by a cost function, a cost value based on the generated corresponding class label and the corresponding probability score (508);
generating one or more gradient maps using the cost value and the generated one or more corresponding convolutional feature maps (510);
generating one or more target activation maps based on the generated one or more gradient maps and the generated one or more corresponding convolutional feature maps (512);
and
identifying a tentative location of the target object of interest in the video frame based on the one or more target activation maps (514).

5 more elements

A system (100) comprising:

a memory (102) storing instructions;
one or more communication interfaces (106);
and
one or more hardware processors (104) coupled to the memory (102) via the one or more communication interfaces (106), wherein the one or more hardware processors (104) are configured by the instructions to:
receive, by a neural network (202), a first input comprising a video frame, and a second input comprising a target object of interest in the video frame, wherein a corresponding class label is generated based on the second input;
receive, at a first part (204) of the neural network (202), a current frame to generate one or more corresponding convolutional feature maps;
generate, by using a softmax layer (208) of the neural network, a probability score for the generated one or more corresponding convolutional feature maps;
generate by a cost function (210) of the neural network (202), a cost value based on the generated corresponding class label and the corresponding probability score;
generate one or more gradient maps using the cost value and the generated one or more corresponding convolutional feature maps;
generate, by using a target activation map generator (212), one or more target activation maps based on the generated one or more gradient maps and the generated one or more corresponding convolutional feature maps;
and
identify a tentative location of the target object of interest in the video frame based on the one or more target activation maps.

9 more elements

One or more non-transitory machine readable information storage mediums comprising one or more instructions which when executed by one or more hardware processors cause:

receiving, at a neural network, a first input comprising a video frame, and a second input comprising a target object of interest in the video frame, wherein a corresponding class label is generated based on the second input;
receiving, at a first part of the neural network, a current frame to generate one or more corresponding convolutional feature maps;
generating, by a softmax layer of the neural network, a probability score for the generated one or more corresponding convolutional feature maps;
generating by a cost function, a cost value based on the generated corresponding class label and the corresponding probability score;
generating one or more gradient maps using the cost value and the generated one or more corresponding convolutional feature maps;
generating one or more target activation maps based on the generated one or more gradient maps and the generated one or more corresponding convolutional feature maps;
and
identifying a tentative location of the target object of interest in the video frame based on the one or more target activation maps.

5 more elements

Dependencies are inferred from the claim text.

Neural network based position estimation of target object of interest in video frames

Life of this patent

Neural network based position estimation of target object of interest in video frames

Life of this patent

Abstract

Claims18