This project is a real-time Emotion Detection system with an intelligent chatbot assistant. It captures video input from the user, analyzes emotions periodically, and provides personalized responses or stress-reducing recommendations using Google's Gemini API.
- 🎥 Real-time Webcam-based Emotion Detection
- 🧠 Emotion classification using pre-trained ML models
- 🤖 Chatbot integration for contextual responses
- 💬 Gemini-powered suggestions for:
- Stress reduction
- Song or activity recommendations
- 📸 Periodic screenshot capture (every 3–5 seconds)
- 📡 FastAPI backend + Next.js frontend
- 🔐
.env-based API key management
| Area | Tools/Frameworks |
|---|---|
| Frontend | Next.js (React + JS) |
| Backend | FastAPI |
| ML/Emotion | Python, OpenCV |
| Chatbot API | Google Gemini API |
| Styling | Tailwind CSS |
- User visits frontend and allows webcam access.
- Video is captured, and frames are periodically sent to the backend for emotion analysis.
- Emotion is detected from the video frames and passed to the chatbot API for personalized interaction.
- Gemini responds with personalized suggestions based on the detected emotion (e.g., stress reduction techniques or song/activity recommendations).
- UI updates with the chatbot's feedback, providing the user with the suggested response.
- 🎯 Audio-based emotion detection
- 📊 Dashboard for emotion trends
- 🧬 Sentiment analysis from text input
- 🌐 Multi-language support
- Shreya Nithin
- Sidharth V S
- Vivek V Joseph
- Zahaan Sajid