A Next.js web application for memory support with face recognition.
- Person Database: Add and manage people with photos, relationships, key facts, and recent topics
- Camera & Face Recognition: Real-time face detection and matching against saved people
- Speech Recognition: Convert speech to text and track conversations
- Accessibility: Large text mode and high contrast mode for easier use
- Conversation Tracking: Save and review past conversations
- Node.js 18+ installed
- npm or yarn
- Install dependencies:
npm install-
Download face-api.js models: Run the provided script to download all required models:
./scripts/download-models.sh
Or manually download them from face-api.js models and place them in
public/models/:tiny_face_detector_model-weights_manifest.jsontiny_face_detector_model-shard1face_landmark_68_model-weights_manifest.jsonface_landmark_68_model-shard1face_recognition_model-weights_manifest.jsonface_recognition_model-shard1
-
Run the development server:
npm run dev- Open http://localhost:3000 in your browser.
- Add People: Go to the "People" page and add people with their photos and information
- Use Camera: Go to the "Camera" page, start the camera, and it will automatically detect and match faces
- Track Conversations: Enable speech recognition to track what's being said
- Customize: Adjust settings for large text and high contrast modes
- Camera access (required for face recognition)
- Microphone access (required for speech recognition)
- Modern browser with Web Speech API support (Chrome, Edge, Safari)
- Data is currently stored in localStorage (browser storage)
- Face recognition requires good lighting and clear view of faces
- Speech recognition works best in quiet environments
- MongoDB integration for persistent storage
- OpenAI integration for conversation summarization
- Cloud storage for images
- Conversation history page
- Caregiver dashboard