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

Gesture recognition method based on attention guidance airspace graph convolution simple loop unit

A simple loop and gesture recognition technology, applied in the field of pattern recognition and computer vision, can solve the problems of high recognition calculation cost and limited gesture recognition applications

Pending Publication Date: 2022-03-18
HANGZHOU DIANZI UNIV
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In many fine dynamic gestures, spatio-temporal graph convolution can still show the spatial and temporal characteristics of gestures, but it inevitably brings too high recognition calculation costs, which limits the application of gesture recognition in high real-time human-computer interaction scenarios. application

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
  • Gesture recognition method based on attention guidance airspace graph convolution simple loop unit
  • Gesture recognition method based on attention guidance airspace graph convolution simple loop unit
  • Gesture recognition method based on attention guidance airspace graph convolution simple loop unit

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0030] The following describes in detail the present invention's dynamic gesture recognition method based on attention-guided spatial graph convolution simple loop unit in conjunction with the accompanying drawings. figure 1 for the implementation flow chart.

[0031] Such as figure 1 , the implementation of the method of the present invention mainly includes four steps: (1) input the gesture data after feature fusion enhancement into the spatial graph convolution SRU network (SGC-SRU); (2) introduce the knuckle attention guidance module to obtain attention Guide the spatial graph convolutional SRU network (ASGC-SRU); (3) introduce the attention-enhanced spatial graph discarding mechanism (ASD) (4) finally aggregate the features and recognize the dynamic gesture category.

[0032] Each step will be described in detail below one by one.

[0033] Step 1: Embed the gesture airspace map convolution operator into a simple recurrent unit, construct a simple recurrent unit for spat...

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 gesture recognition method based on a convolution simple loop unit of an attention guidance airspace graph. The method comprises the following steps of: firstly, embedding a spatial domain graph into a gate structure of a simple cycle unit (SRU) in a convolution manner, so that the SRU with high-speed parallel computing capability can model time domain and spatial domain information of a complex gesture; secondly, a knuckle attention guiding module is introduced, so that more important knuckles have higher attention. Then, a regularization method for discarding the attention enhancement graph is introduced, and the defect of gesture data overfitting is relieved. According to the method, for fine dynamic gestures, high recognition accuracy is achieved, meanwhile, excellent recognition calculation efficiency is kept, and the method has wide application prospects in the fields of man-machine interaction, virtual reality, sign language translation and the like.

Description

technical field [0001] The invention belongs to the fields of computer vision and pattern recognition, uses attention-guided airspace map convolution simple loop unit to perform spatio-temporal modeling of dynamic gestures represented by joint points, and combines attention-enhanced regularization method of airspace map discarding to alleviate overfitting problem, implementing dynamic gesture classification. Background technique [0002] Gesture recognition, as an intuitive and effective interaction tool, plays an important role in many interactive experiences and is a mainstream interaction method. Its applications include sign language translation, virtual reality, and assisted living. Gestures can be divided into static gestures and dynamic gestures. Static gesture recognition is to recognize gestures from a single image, so it runs faster, but the application scenarios have greater limitations. Dynamic gesture recognition, on the other hand, recognizes gestures from a ...

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
IPC IPC(8): G06V40/20G06V10/764G06V10/80G06V10/82G06K9/62G06N3/04G06N3/08
CPCG06N3/084G06N3/048G06N3/045G06F18/2415G06F18/253
Inventor 佘青山陈炫琦武薇高云园范影乐
Owner HANGZHOU DIANZI UNIV