Hybrid Token Efficient Gestor Detector

IshaaraBol is an AI-powered web application designed to bridge the communication gap between sign language users and people who do not understand sign language. Using real-time computer vision and artificial intelligence, the system recognizes hand gestures through a webcam and translates them into meaningful text and speech, enabling seamless and accessible communication. The application supports both single-hand and two-hand gesture recognition, allowing it to understand more complex signs with improved accuracy. Each gesture is detected continuously for a few seconds before being confirmed, reducing false detections and increasing reliability. To provide intelligent interaction beyond gesture recognition, the project integrates the Fireworks AI API. Once a gesture or phrase is recognized, Fireworks AI generates natural language responses, explanations, or conversational assistance, transforming the system from a simple gesture detector into an interactive AI assistant. The platform also includes a gesture sample collection module, where users can capture and securely store gesture images. These samples can be used to expand the dataset, improve AI model accuracy, and support future training and research. An administrative dashboard allows gesture data to be managed efficiently while maintaining organized records. Built with a modern web architecture, the frontend provides a responsive interface for live camera streaming, gesture visualization, and AI responses, while the backend handles gesture processing, API communication, data storage, and user management. The system is designed to be scalable, enabling future enhancements such as multilingual translation, custom gesture creation, mobile deployment, and offline AI inference on AMD-powered hardware.