Supercharge Your Innovation With Domain-Expert AI Agents!

Human body behavior recognition method based on deep ConvLSTM and double-flow fusion network

A technology that integrates networks and recognition methods. It is applied in the field of human behavior recognition based on deep ConvLSTM and dual-stream fusion networks. It can solve the problems of destroying spatial information and low accuracy of behavior recognition. Fitting effect

Pending Publication Date: 2022-05-13
XIAN UNIV OF POSTS & TELECOMM
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The present invention provides a human behavior recognition method (TS-ConvLSTM) based on deep ConvLSTM and dual-stream fusion network to overcome the problem of only focusing on short-term action information in the prior art, introducing RNN to destroy spatial information, and the accuracy of behavior recognition is not high. question

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
  • Human body behavior recognition method based on deep ConvLSTM and double-flow fusion network
  • Human body behavior recognition method based on deep ConvLSTM and double-flow fusion network
  • Human body behavior recognition method based on deep ConvLSTM and double-flow fusion network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0024] The invention is further described below in combination with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical scheme of the invention, and cannot limit the protection scope of the invention.

[0025] as Figure 1 As shown, the human behavior recognition method based on deep convlstm and dual stream fusion network provided by the invention comprises the following steps:

[0026] Step 1: use OpenCV to extract video frames and corresponding optical flow feature images, select one optical flow image every 3 frames, and stack 10 dual channel optical flow images into a new 20 channel frame;

[0027] Step 2: a one-way connection from time flow to space flow is adopted to connect the same layer of the two flows directly, so as to strengthen the representation ability of similar actions, so that the spatiotemporal network can interact;

[0028] Step 3: sum is used to fuse spatiotemporal features. In the same position and corr...

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 human body behavior recognition method based on a deep ConvLSTM and a double-flow fusion network, and belongs to the field of computer vision and mode recognition. According to the technical scheme, the method comprises the following steps: 1, extracting a video frame and a corresponding optical flow feature map by adopting OpenCV, receiving an original video frame by a spatial flow network, and receiving an optical flow frame as input by a time flow network; 2, double-flow CNN feature extraction; 3, performing feature fusion by adopting a Sum fusion strategy; and 4, performing global average pooling on the feature description obtained in the step 3 in an output state of ConvLSTM, and applying the global average pooling to a softmax layer to complete feature sequence classification. According to the method and the device, the problem of inaccurate identification caused by incapability of accurately acquiring comprehensive behavior action information in the prior art is solved, the capability of capturing long-term video time information and the capability of comprehensively acquiring the behavior action information are improved, and the accuracy of video behavior identification is further improved.

Description

technical field [0001] The invention belongs to the field of computer vision and pattern recognition, in particular to a human behavior recognition method based on deep convlstm and dual stream fusion network. Background technology [0002] With the explosive growth of video data, human behavior recognition in video has become an important task. Human behavior recognition is an important research direction of using computer vision technology to process video content. Human behavior recognition is an important part of video intelligent analysis. Its purpose is to analyze and understand the video sequence by computer, and then identify the human actions in the video. At present, human behavior recognition technology has important research and application value in the fields of security, intelligent medical treatment, human-computer interaction and video retrieval. [0003] For video, the research of human behavior recognition mainly includes three core parts: data preprocessing, hu...

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): G06V40/20G06K9/62G06N3/04G06N3/08G06V10/40G06V10/80G06V10/82
CPCG06N3/084G06N3/044G06N3/045G06F18/253
Inventor 潘晓英李守坤薛玉峰
Owner XIAN UNIV OF POSTS & TELECOMM
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com