AI Assistance tool for Sign language using Computer Vision
Humans communicate with one another to share their thoughts, feelings and experiences. But this is not the case for the deaf-mute. There are many people who face these disabilities right from birth or because of issues such as Selective mutism, Aphasia syndrome, Alport syndrome, Norrie syndrome etc.A large number of the population is disconnected from the mainstream hearing-dominated society and lie at the risk of being marginalized because people who are limited to using only speech can’t communicate with them. A lack of accessibility to support the conversation between both communities also adds to the problem. Due to this , there is a communication gap between the deaf-mute and those who can speak.To bridge this gap, a non-verbal language called sign language exists. According to the World Federation of the Deaf, there are more than 70 million deaf people worldwide. More than 80% of them live in developing countries. Collectively, they use more than 300 different sign languages. Sign languages are fully fledged natural languages, structurally distinct from spoken languages.In this project, we have implemented a solution for INDIAN SIGN LANGUAGE.
- Our solution to this problem is an AI assistance for sign language tool that works based on computer vision and machine learning using technologies like OpenCV,Mediapipe,BERT,Streamlit.etc
- The tool gets the sign language gesture performed by the person as input from the camera and converts that into speech.SImilarly , it converts text to Indian sign language gestures.
- We’ve trained the ML model using the SVM algorithm.
- We have also added features like word autocomplete,Next word prediction and Backspacing
It is recommended to use a Virtual Environment to run this project
Install all the required packages using
pip install -r requirements.txt
Open terminal in the directory where the home.py file is present and run the command
streamlit run Home.py
Feel free to contact us on LinkedIn - Harikrishnan S , Rohit Arrunachalam , Isha.
Make a pull request on this repo if you would like work towards improving this project.