MindMate: A Mental Health Support Chatbot with LLM Integration and Emotion Detection

This research presents an integrated system that combines a mental health chatbot and an emotion recognition module to deliver empathetic, personalized digital mental health-care. Built using FastAPI, LLaMA-2-7B, FAISS, and Hugging Face embeddings, the chatbot provides concise, context-aware responses and directs users to professional resources when sensitive topics are detected. The emotion recognition module, developed using Keras and TensorFlow, employs a convolutional neural network (CNN) to classify emotions such as happiness, sadness, anger and fear through facial expression analysis. By integrating real time emotional insights into the chatbot’s response mechanism, the system enhances user interaction, ensuring both contextual accuracy and empathetic support. This solution highlights the potential of AI-driven technologies in addressing gaps in mental health services.