THE ISHARA-LIPI DATASET

Of Bangla Sign Language Digits and Characters


Md. Sanzidul Islam , Core Member, DIU NLP Lab
Sadia Sharmin Mousumi, Member, DIU NLP Lab
Nazmul A. Jessan, DIU
Syed Akhter Hossain, Head, Dept. of CSE, DIU


Data Number of class Sets per class Total number of data
Digits 10 100 1000
Vowels 6 50 300
Consonents 30 50 1500
Digits: 1000 Images
Characters: 1800 Images
Total: 2800 Images

Bangla language carries a bloody history after our historical language step in the year of 1952. But what about the language of deaf community in Bangladesh? To find this answer, we were aimed to work for decreasing the gap between hearing impaired people and generals. Ishara-Lipi, the first multipurpose comprehensive open access isolated dataset for Bangla Sign Language(BdSL) Digits and Characters. It will help the researchers to work with sign recognizer, machine translator and to develop the aiding tools.


Data Total Scopes
Digits 1000 Images View Download Publication Cite
Characters 1800 Images View Download Publication Cite

It's too difficult to collect hand sign's image data in Bangladesh. Because the deaf community and collaborators aren't concern enough about modern technologies and approaches. We, the tem Ishara-Lipi collected hand signs data from different deaf community, institutes and university volunteers.

The whole dataset is devided into two portions- one is for digits(1, 2, 3, . . ., 9) and another is for characters(1, 2, 3, . . ., 36). The digits dataset contains 100 sets of 10 Bangla basic sign digits (0, 1, 2 . . ., 9), collected from different deaf and general volunteers from different institutes. After discarding maximum errors and performing different preprocessing methods, 1000 images of Bangla sign language isolated digits were included in the final dataset. After collectiong raw data, some effective preprocessing methods were performed for making those data useable for computer vision or any other application development purposes.

And the characters dataset contains 50 sets of 36 Bangla basic sign characters, collected by the help of different deaf and general volunteers from different institutes. 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 signs.


Publications List-

  1. Islam, Sanzidul, et al. "A Potent Model to Recognize Bangla Sign Language Digits Using Convolutional Neural Network." Procedia computer science 143 (2018): 611-618.
  2. Islam, Md Sanzidul, et al. "Ishara-Lipi: The First Complete MultipurposeOpen Access Dataset of Isolated Characters for Bangla Sign Language." 2018 International Conference on Bangla Speech and Language Processing (ICBSLP). IEEE, 2018.
  3. Ishara-Bochon: The First Multipurpose Open Access Dataset for Bangla Sign Language Isolated Digits. [Communications in Computer and Information Science (CCIS), Springer, 2018], Waiting for publishing.
  4. A Simple and Mighty Arrowhead Detection Technique of Bangla Sign Language Characters with CNN. [Communications in Computer and Information Science (CCIS), Springer, 2018], Waiting for publishing.

Terms of Use-

The annotations in this dataset along with this website belong to the Ishara-Lipi Consortium and are licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

Creative Commons License

Images

The Ishara-Lipi Consortium does own the copyright of the images. The users of the images accept full responsibility for the use of the dataset.

Software

Copyright (c) 2019, Ishara-Lipi Consortium. All rights reserved. Redistribution and use software in source and binary form, with or without modification, are permitted provided that the following conditions are met:

THIS SOFTWARE AND ANNOTATIONS ARE PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS AS IS AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.


Other Collaborators-

Shahariar Azad Rabby, Core Member, DIU NLP Lab
Sheikh Abujar, Supervisor, DIU NLP Lab

Special Thanks to-

Centre for Disability in Development (CDD), Savar
Mirpur Deaf School, Dhaka
Bijoynagar Deaf School, Dhaka
Mymensing Deaf School, Mymensing

©2019 Team ISHARA-LIPI. All Rights Reserved.