Feasible LSTM Model for Detection of Sign and Body Language

Abstract

The use of sign and body language facilitates communication between hearing individuals and those with hearing loss. They play an important role in facilitating effective human-human and human-computer interactions. It is a visual language composed of hand gestures and facial expressions to identify the meaning conveyed by the signer. Body language detection is the process of analysing the nonverbal cues and gestures used in communication to understand the emotional state and intentions of the speaker. Deep learning methods have recently demonstrated promising results in a range of computer vision tasks, including gesture detection. Using a media pipe holistically, extraction of essential points from the body, hands, and face is possible. TensorFlow and Keras are further utilized to construct Long Short-Term Memory LSTM feasible models that can predict on-screen behaviour presenting an innovative approach for sign and body language detection using LSTM.