Unlock instant, AI-driven research and patent intelligence for your innovation.

Graph Binary Feature Learning Method and Device Based on Deep Reinforcement Learning

A technology of reinforcement learning and feature learning, applied in the field of computer vision and machine learning.

Active Publication Date: 2020-09-25
TSINGHUA UNIV
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the binary feature learning method in the related art does not consider the relationship between the bits of the feature, and each bit of the independent learning feature may lead to feature bits located at the boundary of the binarization, and these bits are easily affected by noise. Therefore, the robustness of the feature is not strong, and it needs to be solved

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Graph Binary Feature Learning Method and Device Based on Deep Reinforcement Learning
  • Graph Binary Feature Learning Method and Device Based on Deep Reinforcement Learning
  • Graph Binary Feature Learning Method and Device Based on Deep Reinforcement Learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0038] Embodiments of the present invention are described in detail below, examples of which are shown in the drawings, wherein the same or similar reference numerals designate the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the figures are exemplary and are intended to explain the present invention and should not be construed as limiting the present invention.

[0039] The following describes the method based on deep reinforcement learning according to the embodiment of the present invention with reference to the drawings Figure II value feature learning method and device, firstly, the deep reinforcement learning based Figure II value feature learning method.

[0040] figure 1 It is the deep reinforcement learning based on the embodiment of the present invention Figure II value feature learning method

[0041] Such as figure 1 As shown, the deep reinforcement learning-based Figur...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a method and device for learning binary features of images based on deep reinforcement learning, wherein the method includes: extracting deep real-valued features of images; and obtaining the basic quantity of the mining network for the relationship between bits according to the relationship between bits of deep reinforcement learning , where the basic quantities include state, transition matrix, actions and rewards to train the inter-bit relationship mining network; extract robust features through the inter-bit relationship mining network and the feature extraction network of mutual information. This method can extract robust features through inter-bit relationship mining network and mutual information feature extraction network, and effectively improve feature robustness.

Description

technical field [0001] The present invention relates to the technical fields of computer vision and machine learning, in particular to a deep reinforcement learning-based Figure II Value feature learning method and device. Background technique [0002] One of the very important basic problems in the field of computer vision is the visual recognition task, which is widely used in various visual tasks, such as general object recognition, scene recognition, face recognition and fingerprint recognition. The visual recognition task is essentially a pattern recognition task, which has the characteristics of high data dimensionality, large amount of data, and large data differences. The basis and basis of visual recognition tasks are visual features, which refer to feature vectors corresponding to pictures. A "good" feature satisfies the characteristics that the similarity of the feature vectors of the same kind of pictures is strong, and the similarity of the feature vectors of ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06K9/38G06K9/46G06K9/62G06N3/04
CPCG06V10/28G06V10/462G06N3/045G06F18/24G06F18/214
Inventor 鲁继文周杰段岳圻
Owner TSINGHUA UNIV