Bidirectional Indian Sign Language Translation: A Survey of Integrating Context, Emotion, and Cloud APIs like AWS Polly

This survey analyzes the landscape of bidirectional translation between Indian Sign Language (ISL) and multiple Indian regional languages, which addressed critical limitations in existing technologies such as unidirectionality, context blindness, emotional deafness, and linguistic exclusion, failing to meet the communication needs of India’s diverse Deaf community. This paper synthesize research across five modules: (1) Input Processing for real-time sign capture and speech/text input, (2) ISL Recognition using MediaPipe and deep learning, (3) Context & Emotion Intelligence Layer, (4) Text-to-ISL Generation with emotionally expressive AR-based 3D Signing Avatar, and (5) Multilingual Output using AWS Polly for emotionally infused speech synthesis in Hindi, Tamil, and Marathi. This survey analysis is a model shift from treating sign language translation as purely syntactic conversion to a universal process conveying verbal content, emotional variation, and cultural context