Machine Learning based ISL Identification and Translation

Sign language is an essential means of communication for the deaf and hard-of-hearing community. However, effective communication between sign language users and those unfamiliar with sign language can be challenging. The primary goal is to utilize the machine learning to automatically identify sign language gestures and translate them into easily understandable formats. This research presents a comprehensive sign language detection system that captures sign language gestures, detects them, and provides output in text using LSTM (Long Short-Term Memory) and Transformers with an accuracy of 79%. This multimodal approach ensures the system helps in understanding the sign language.