Sign Language Interpretation Employing SVM and CNN - Comparative Analysis
Arpita Hosur, Patadayya Chikkamath, Gagan Muddebihal, Rakshita Aramani, Gujanatti Rudrappa · 2025
The Sign Language Interpreter is a crucial tool for deaf and mute individuals, but many people struggle to understand sign language. This paper proposes a solution using machine learning models for gesture recognition, specifically CNN and SVM. The system uses 26,000 images of ASL to evaluate their performance, with the CNN model achieving 93.83 % accuracy. The accuracy of CNN model achieved was 93.83 %. This approach enhances independence and inclusiveness for the speechless, allowing real-time gesture recognition and translation. The system reduces the communication gap by fostering real time gesture recognition and translation which in turn enhances the independence of the speech impaired. The system outperforms traditional methodologies in scenarios requiring both interpretation and real-time scalability, making it superior to traditional methods in situations requiring interpretation and real-time scalability.