Hand Sign Language Detection Using Deep Learning and Computer Vision
Sujata Joshi, Bhaskar Sharma, Dharmender, Divyansh Tripathi, Akshat Shah
Abstract
This work presents an innovative system/app designed to assist individuals with visual or hearing impairments in communicating with the outside world using American Sign Language (ASL). By leveraging machine learning and image processing techniques, the proposed system captures and processes hand gestures, enabling the optimal transliteration of sign language alphabets and numbers into machine-readable English text. The system employs deep learning or random forest algorithms to classify the gestures, followed by mapping them to specific alphabets or numbers. The resulting text is presented audibly, bridging the communication gap and facilitating interactions between ASL users and those unfamiliar with sign language. Through the integration of technology, this work aims to enhance inclusivity and accessibility for individuals having disabilities, fostering greater engagement and interaction with the wider world.