Scopus Indexed Publications

Paper Details


Title
Ishara-Lipi: The First Complete MultipurposeOpen Access Dataset of Isolated Characters for Bangla Sign Language
Author
Md. Sanzidul Islam, A K M Shahariar Azad Rabby, Nazmul Ahsan Jessan, Sadia Sultana Sharmin Mousumi, Syed Akhter Hossain,
Email
sanzid.swe@diu.edu.bd
Abstract
Collecting hand gesture data for sign language is too much difficult to researchers. Ishara-Lipi, the first complete isolated characters dataset of Bangla Sign Language (BdSL) is conducted in this article. It will help to increase interaction between hearing impaired community and general people. The dataset contains 50 sets of 36 Bangla basic sign characters, collected by the help of different deaf and general volunteers. In Bangla Sign Language sign characters there have 6 vowels and 30 consonants by which they can finger spell all Bangla words. In Ishara-Lipi dataset, after discarding mistakes and preprocessing, 1800 character images of Bangla Sign Language were included in the final state. This dataset could be used to develop computer vision based or any kind of system that approves users to search the meaning of BdSL sign.

Keywords
Hangla Sign Language , Computer vision , Upen Source , Sign language characters , BdSL , NLP , Machine Learning , Pattern Recognition , Sign Language Dataset
Journal or Conference Name
International Conference on Bangla Speech and Language Processing, ICBSLP 2018
Publication Year
2018
Indexing
scopus