ASL Recognition System
American Sign Language Recnognition system in real time, using trained Neural Network
Course ECS 271: Machine Learning & Discovery (UC Davis)
Instructor Prof. Hamed Pirsiavash
Team Muhammad Hassnain, Nafiz Imtiaz Khan, Md Rian Nabil Latif
Quarter Fall 2024
Sign Language is a vital mode of communication around the world for individuals who are deaf or hard of hearing. Among various such languages American Sign Language (ASL) stands as one of the most extensively developed systems. However, a major problem still remains, it is only accessible to inidividuals who have learned it. This limitartion creates a siginificant gap in communication between the users of ASL and the broader population. In everyday scenarios, such as shopping, healthcare or education, this communication barrier can lead to exclusion and dependency of interpreters, who are not always available. To address this, we envision a system that automatically translates sign language into a text or speech in real time, enabling individuals to communicate without requiring prior knowledge of sign language. This project is however a proof of concept and is only limited to english alphabets. Here is a collage of pictures showing model real time detection capabilies. Please ignore my tired face, these were taken at 4 am.
One day, I will deploy the model so that you can try it out yourself.