Sign language parameters classification from 3D virtual charactarers

Deaf and hard of hearing individuals are facing lot of barriers that prevent them from accessing to information. Signing avatars help them to overcome these barriers. These virtual characters are able to “speak” Sign language and subsequently able to translate any kind of information into Sign language. Recently, thanks to the advances in virtual reality and human modeling techniques, signing avatars are increasingly used by deaf communities. Moreover, thanks to the apparition of new standards, 3D signing avatars are constantly exchanged and uploaded to the World Wide Web. Unfortunately, current search engines and catalog systems that deal with signing avatars are not indexing them efficiently. In this paper, we present a new approach to recognize and index 3D signed contents based on the recognition and classification of sign language parameters. Our approach uses an adaptation of the Longest common subsequence algorithm combined with Minkowski similarity measures.