International Journal of Innovative Research in Computer and Communication Engineering
ISSN Approved Journal | Impact factor: 8.771 | ESTD: 2013 | Follows UGC CARE Journal Norms and Guidelines
| Monthly, Peer-Reviewed, Refereed, Scholarly, Multidisciplinary and Open Access Journal | High Impact Factor 8.771 (Calculated by Google Scholar and Semantic Scholar | AI-Powered Research Tool | Indexing in all Major Database & Metadata, Citation Generator | Digital Object Identifier (DOI) |
| TITLE | SignSpeak AI: Real-Time Sign Language Recognition using Hybrid Deep Learning and Explainable AI |
|---|---|
| ABSTRACT | Communication barriers between the deaf/hard-of-hearing community and the hearing population persist due to limited awareness and the scarcity of accessible sign language interpretation services, while traditional sign-language recognition systems are restricted by limited datasets, static-only gesture detection, and lack of transparency in deep learning models. This paper presents SignSpeak AI, a real-time hybrid gesture-to-speech translation system integrating Convolutional Neural Networks (CNNs) for static sign recognition, LSTM/Transformer architectures for dynamic gesture interpretation, and Explainable AI (XAI) methods such as Grad-CAM and SHAP to ensure model transparency and trust. The system operates using standard webcams, processes input entirely in real time, and features a sentence-generation component that converts detected gestures into meaningful English speech. Experimental evaluation across alphabets (A–Z), numbers (0–9), and 25 commonly used sign-language terms demonstrates high prediction accuracy, low-latency inference, and strong interpretability, making SignSpeak AI a scalable, accessible, and transparent assistive solution suitable for educational, clinical, and public communication environments. |
| AUTHOR | DR. NAVEEN H M, AKASH G C, B M DEVIKA, IRAMMA SHANKRAPPA GUMAGANDI, SHREEKANTH K UTTARKAR Assistant Professor, Department of Computer Science and Engineering, Bapuji Institute of Engineering and Technology, Davangere, Karnataka, India UG Student, Department of Computer Science and Engineering, Bapuji Institute of Engineering and Technology, Davangere, Karnataka, India |
| VOLUME | 177 |
| DOI | DOI: 10.15680/IJIRCCE.2025.1312004 |
| pdf/4_SignSpeak AI Real-Time Sign Language Recognition using Hybrid Deep Learning and Explainable AI.pdf | |
| KEYWORDS |
