A Skeleton Gesture Recognition Method Based on Bidirectional Independent Recurrent Neural Network

A cyclic neural network and gesture recognition technology, which is applied in the field of skeleton gesture recognition based on bidirectional independent cyclic neural network, can solve the problems of gradient explosion of cyclic structure, inability to obtain high accuracy of skeleton gesture recognition tasks, etc., and achieves high accuracy. Effect

Active Publication Date: 2020-04-24
UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF2 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, it is still affected by the gradient explosion and gradient disappearance of the loop structure, and cannot achieve high accuracy in the skeleton gesture recognition task.

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 Skeleton Gesture Recognition Method Based on Bidirectional Independent Recurrent Neural Network
  • A Skeleton Gesture Recognition Method Based on Bidirectional Independent Recurrent Neural Network
  • A Skeleton Gesture Recognition Method Based on Bidirectional Independent Recurrent Neural Network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0031] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

[0032] A kind of skeleton gesture recognition method based on two-way independent recurrent neural network, comprises the following steps:

[0033] Step 1: Obtain the skeleton gesture dataset and perform preprocessing;

[0034]The DHG data set, a general-purpose skeleton gesture data set, is used to train the neural network; the DHG data set contains 2800 time series of 14 gestures, including grabbing (G), tapping (T), zooming in (E), twisting (P), Rotate clockwise (RC), Rotate counterclockwise (RCC), Swipe right (SR), Swipe left (SL), Swipe (SU), Swipe down (SD), Swipe diagonally (SX ), sliding + (S+), v-shaped sliding (SV), and swinging (SH), a total of fourteen gestures; when distinguishing single-finger gestures and multi-finger gestures, they can be divided into 28 gesture categories; according to the framework based on deep learning The...

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 skeleton gesture recognition method based on a two-way independent cyclic neural network, which comprises the following steps: Step 1: Obtain a skeleton gesture data set and perform preprocessing; Step 2: Construct a bidirectional independent cyclic neural network; Step 3: Combine the steps The data set obtained in 1 is input into the neural network constructed in step 2 for training; step 4: the neural network obtained in step 3 is used for recognition of skeleton gestures; the present invention has the ability to extract advanced spatial information and time-related information and can By extracting two-way time-related information, the accuracy of skeleton gesture recognition is high.

Description

technical field [0001] The invention relates to a skeleton gesture recognition method, in particular to a skeleton gesture recognition method based on a bidirectional independent cyclic neural network. Background technique [0002] Gesture recognition technology is widely used in human-computer interaction fields such as virtual reality, sign language recognition, and robot control; with the development of non-wearable depth sensors such as Microsoft Kinect and Intel RealSence, gesture recognition algorithms based on skeleton data are widely used. However, due to the small spatial difference, the skeleton gesture recognition task has high requirements for time-related information, so it has high requirements for the algorithm's ability to extract spatial and temporal information. Skeleton gesture recognition algorithms can be divided into gesture recognition algorithms based on artificial features and gesture recognition algorithms based on deep learning; the method based on...

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): G06K9/00G06N3/04
Inventor 李帅朱策郑龙飞张铁高艳博
Owner UNIV OF ELECTRONICS SCI & TECH OF CHINA
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap