Mostakim, MoinShakil, ArifAlam, Md. Golam RabiulSelim, Bushra BinteIqbal, MalihaShahriar, AsifFaria, FauziaMostafa, Rafid2021-09-042021-09-0420212021ID 21141052ID 21141050ID 16301040ID 17141007ID 16101069http://hdl.handle.net/10361/14969Cataloged from PDF version of thesis.Includes bibliographical references (pages 41-44).This thesis is submitted in partial fulfillment of the requirements for the degree of Bachelor of Science in Computer Science and Engineering, 2021.Sign gesture, which is one type of non-audible specialized strategy is the medium to correspond with individuals having auditory and talking incompetency. There are numerous computerized methods of creating gesture-based communication to provide aid among the hearing impaired. Particularly, for Bengali sign dialect, quite a few measures have been taken for generation of automated Bangla sign gestures. With an authentic dataset and approach, an apparent communication mode to assist this non-privileged community can be attained. Our method proposes a convolutional neural network (CNN) to derive a picture of the appropriate sign gesture of a particular Bangla alphabet. After our examinations and multiple experiments, we have come up with the simplest and most striking methodology to perform the mentioned task. Our model worked promptly and provided remarkable accuracy. Needless to mention that, communication through gestures aided by artificial means is another corner that needs to be explored more. Henceforth, our work can have an added value to this ongoing inspection.44 pagesenBrac University theses are protected by copyright. They may be viewed from this source for any purpose, but reproduction or distribution in any format is prohibited without written permission.Sign Language GenerationNeural NetworkNeural Networks for Sign LanguageGenerative SignCNNInceptionv3Sign language.Thesis