Gesture recognition method based on deep learning

Tong Du, Xuemei Ren, Huichao Li

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

15 Citations (Scopus)

Abstract

With the rapid development of science and technology, human-computer interaction is born more frequently around us. Human motion analysis and recognition based on attitude sensor is a new field, which overcomes many shortcomings and limitations of motion recognition based on video and is more practical. In this paper, we proposes a new method based on time gesture recognition. By analyzing the kinematics of gestures, the features of gestures are extracted and classified using Recurrent Neural Networks and their variant networks. The methods achieved an accuracy of over 98% in 16 experimenters. The results show that the algorithm can quickly and accurately identify gestures.

Original languageEnglish
Title of host publicationProceedings - 2018 33rd Youth Academic Annual Conference of Chinese Association of Automation, YAC 2018
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages782-787
Number of pages6
ISBN (Electronic)9781538672556
DOIs
Publication statusPublished - 6 Jul 2018
Event33rd Youth Academic Annual Conference of Chinese Association of Automation, YAC 2018 - Nanjing, China
Duration: 18 May 201820 May 2018

Publication series

NameProceedings - 2018 33rd Youth Academic Annual Conference of Chinese Association of Automation, YAC 2018

Conference

Conference33rd Youth Academic Annual Conference of Chinese Association of Automation, YAC 2018
Country/TerritoryChina
CityNanjing
Period18/05/1820/05/18

Keywords

  • Attitude sensor
  • Gesture recognition
  • Recurrent Neural Networks

Fingerprint

Dive into the research topics of 'Gesture recognition method based on deep learning'. Together they form a unique fingerprint.

Cite this