Bridging the Communication Gap: An Improved Real-Time Indian Sign Language Translation with Deep Learning

Indian Sign Language (ISL) is the primary mode of communication for the Deaf and Hard-of-Hearing (DHH) community in India. However, communication barriers persist due to limited public awareness of ISL. This work presents a real-time gesture recognition system that translates ISL gestures into text and speech, facilitating interaction between ISL users and non-signers. The system employs Convolutional Neural Networks (CNN) for spatial feature extraction and Long Short-Term Memory (LSTM) networks for capturing temporal patterns in sequential gesture data. A diverse dataset, incorporating variations in hand movements, facial expressions, and body posture, enhances model accuracy. The integration of deep learning techniques allows continuous performance improvement. This solution fosters inclusivity by bridging the communication gap in public spaces, educational institutions, and workplaces. The proposed system offers a scalable, cost-effective approach to improving accessibility and social integration for the DHH community.