Indian Sign Language Recognition System

Communication barriers between the hearing-impaired, speech-impaired, and non-signers create persistent challenges in education, healthcare, and daily interactions. While Indian Sign Language (ISL) is widely used within the deaf community, the lack of public awareness significantly limits effective communication. This paper presents a real-time ISL Recognition System developed using deep learning and computer vision. The system captures hand gestures through a webcam, extracts 21 hand landmarks using Google Mediapipe, and classifies them using a trained Convolutional Neural Network (CNN). The recognized gestures are translated into text and speech for better accessibility. In addition to gesture recognition, the system provides speech-to-text functionality to enable two-way communication between hearing and non-hearing users. The proposed solution is lightweight, cost-effective, and easy to deploy, making it suitable for educational institutions, assistive technologies, and inclusive digital platforms. The results demonstrate high accuracy and robust performance, validating the effectiveness of the proposed framework.