What it does

Mind Compass is an AI-powered mental wellness companion that: Detects your mood in real-time using facial emotion recognition. Recognizes hand gestures and sign language to make interactions more accessible. Enables voice interaction through speech-to-text and text-to-speech for seamless communication. Shows personalized wellness insights, daily trends, and scores on a stylish dashboard. In short, it understands how you feel and guides you toward better mental wellness.

How we built it

Frontend: React + TypeScript + Vite + TailwindCSS + ShadCN UI for a responsive, modern interface. AI/ML: TensorFlow.js for live emotion and gesture detection. Voice Interaction: Web Speech API for STT/TTS, letting users talk to the AI naturally. Animations: Three.js and interactive particles to make the UI dynamic and immersive. Deployment: Hosted on Vercel for instant live demos. We combined multiple modalities (camera, voice, gestures) into a single seamless experience with real-time interactivity.

Challenges we ran into

Optimizing AI models for real-time performance without slowing the app. Ensuring accurate gesture and emotion detection across different lighting and environments. Balancing stylish animations with smooth UX on desktop and mobile. Integrating camera, voice, and 3D UI together without conflicts or lag.

Accomplishments that we're proud of

Built a fully functional live prototype in 36 hours. Integrated voice, camera, and gesture inputs for interactive AI experiences. Created a modern, stylish, and trendy UI that’s responsive and visually appealing. Developed real-time mental health insights with wellness scores and trends.

What we learned

How to combine multiple AI modalities (voice, vision, gestures) in a web app. Real-world application of TensorFlow.js for live browser-based AI. Prompt engineering to generate helpful and engaging AI responses. Optimizing performance while maintaining polished UI/UX with animations.

What's next for Mind Compass

Expand with personalized mental health plans and daily recommendations. Gamify wellness tracking with challenges and streaks to engage users. Enable multi-user and community features for shared experiences. Improve gesture detection accuracy and expand sign language support. Add advanced AI mood prediction based on long-term user trends.

Built With

  • ai
  • frontend
  • vercel
Share this project:

Updates