Gesture to Speech and Text: A Comprehensive Review Proposal on Sign Language Recognition

Gesture to Speech and Text: A Comprehensive Review Proposal on Sign Language Recognition

Mr. Jugal Mistry 1, Mr. Sujal Kabra2, Mr. Rahul Kumawat3, Mr. Hardik Parmar4, Asst. Prof. Ms. Manisha Vasava5

1234Research Scholar, Department of Information Technology, Krishna School Of Emerging Technology & Applied Research, KPGU University, Varnama, Vadodara, Gujarat, India

5Assistant Professor, Department of Information Technology, Krishna School of Emerging

Technology & Applied Research, KPGU University, Varnama, Vadodara, Gujarat, India

---------------------------------------------------------------------***---------------------------------------------------------------------

Abstract - This manuscript presents an extensive review of sign language recognition (SLR) technologies and proposes an innovative hybrid framework to convert gestures into both speech and text. Covering studies from 2014 to 2024, the review details diverse methodologies—including CNN-LSTM networks, Transformer-based models, sensor fusion, and graph convolutional networks—while addressing challenges such as occlusion, hardware dependency, limited datasets, and language specificity. The proposed research focuses on integrating spatial feature extraction, temporal modeling, and attention mechanisms to achieve real-time, multilingual, and emotion-aware SLR. The study also suggests standardization in evaluation and ethical AI practices, aiming to enhance communication accessibility for hearing-impaired communities.

Key Words:  Sign language recognition, gesture-to-speech, text conversion, deep learning, sensor fusion, real-time systems, ethical AI.