Deep Learning System for ISL Recognition Using Long Short-Term Memory Technique

For people with hearing impairments, gesture- and sign-based language is an essential means of communicating ideas. In this research, we offer an ISL sign language recognition system that makes use of deep learning methods. Our approach extracts spatial data from video frames by using a media pipe and a Long Short-term memory (LSTM) to capture the complex hand gestures and movements that are characteristic of sign language. Sequential information and short-term dependencies are modelled using an LSTM network. Some of the most often used ISL indicators were utilized to train the model. The created system recognizes sign language with high accuracy; the findings show that the accuracy is better than previous methods. By promoting inclusion and individual accessibility to communication, this research advances assistive technologies.