Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A pedestrian rerecognition method based on reinforcement learning adaptive partitioning

A pedestrian re-identification and reinforcement learning technology is applied in the field of pedestrian re-identification based on reinforcement learning adaptive block to achieve the effect of improving the recognition accuracy, and enhancing the anti-interference and robustness.

Inactive Publication Date: 2018-12-25
襄阳矩子智能科技有限公司
View PDF3 Cites 16 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] Aiming at the defects and improvement needs of the existing block convolution method, the present invention provides a pedestrian re-identification method based on reinforcement learning adaptive block segmentation. Driven by data, this method adaptively performs segmentation according to the characteristics of the pedestrian image itself. block and extract features for person re-identification

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
  • A pedestrian rerecognition method based on reinforcement learning adaptive partitioning
  • A pedestrian rerecognition method based on reinforcement learning adaptive partitioning
  • A pedestrian rerecognition method based on reinforcement learning adaptive partitioning

Examples

Experimental program
Comparison scheme
Effect test

example

[0110] In order to prove that the pedestrian re-identification method based on reinforcement learning and adaptive segmentation has advantages in both performance and adaptability, the present invention conducts verification and analysis through the following experiments:

[0111] A. Experimental data set

[0112] Dataset 1: Market-1501. The data set is collected at the entrance of a supermarket on the campus of Tsinghua University. It contains a total of 1501 pedestrians with different identities. The shooting equipment is 6 cameras with different angles of view. Model) to generate detection boxes and crop pedestrian targets. In this data set, 12936 pedestrian pictures and 750 identities are divided into the training set, and another 19732 pedestrian pictures and 751 identities are divided into the test set. In addition, there are 3368 pedestrian pictures as the retrieval set, which have the same identity as the test set pictures.

[0113] Dataset 2: DukeMTMC-ReID. This d...

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 belongs to the field of computer vision and pattern recognition, in particular to a pedestrian re-recognition method based on reinforcement learning adaptive partitioning. The inventionintroduces a reinforcement learning method to train an intelligent network, so that the intelligent network can be driven by the data itself, and adaptively decides a partition strategy (the number ofpartitions and the size of partitions) according to the characteristics of the retrieval pedestrian. Experiments show that the invention not only improves the flexibility and generalization ability of the model, but also has a considerable application prospect.

Description

technical field [0001] The invention belongs to the field of computer vision and pattern recognition, and in particular relates to a pedestrian re-identification method based on reinforcement learning adaptive block. Background technique [0002] Against this background, the cameras in the security monitoring system are continuously collecting massive amounts of video image data every day. Often, a single camera has a limited range of monitoring capabilities, and when investigating various cases, it is necessary to search and track across multiple cameras. When a case occurs, it takes a lot of time and manpower to analyze the monitoring data only with police manpower, and it is more likely to miss the best time to solve the case. [0003] Person Re-identification (Re-ID) is an important research direction in computer vision and pattern recognition. The purpose of pedestrian re-identification is to retrieve pedestrian targets that have appeared under different cameras at dif...

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 Applications(China)
IPC IPC(8): G06K9/00G06K9/46G06K9/62
CPCG06V40/10G06V10/50G06F18/24G06F18/214
Inventor 史宇轩
Owner 襄阳矩子智能科技有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com