Turkish Sign Language Recognition With Leap Motion

DEMİRCİOĞLU B., BULBUL G., Köse H.

24th Signal Processing and Communication Application Conference (SIU), Zonguldak, Turkey, 16 - 19 May 2016, pp.589-592 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Doi Number: 10.1109/siu.2016.7495809
  • City: Zonguldak
  • Country: Turkey
  • Page Numbers: pp.589-592
  • Istanbul Technical University Affiliated: Yes

Abstract

The aim of this project is to establish the communication between the hearing impaired people and non-signlanguage-speaking people easier and to raise the awareness of the community about hearing impaired people. In this study, a subset of selected primitive hand shapes of Turkish Sign Language are recognized by a system based on Leap Motion device. The system was tested with selected participants with both right and left hands, then the recognized data is also tested offline with Random Forest and Multi Layer Perceptron methods, and the results were compared. The proposed system requires less data than the machine learning methods, and its success rate is higher.