∗ Developed a real-time ASL recognition model using MediaPipe; trained Random Forest, MLP, XGBoost, and Logistic Regression classifiers to maximize accuracy, reduce prediction time and mitigate overfitting.
∗ Developed a domain-specific Retrieval-Augmented Generation (RAG) LLM chatbot for a smooth human-computer interaction (HCI).
∗ Integration of the SLR system with a domain-specific chatbot to enable real-time webcam-based communication for sign language users without the need for text or voice input.