How manavarya09/restaurant-ai works
Restaurant AI is a computer-vision analytics system for turning CCTV or recorded video into customer-flow and service insights. Its principal runtime path is a video-analysis request to the FastAPI backend, which processes frames with person detection and tracking, then exposes analytics through API routes and a dashboard. Domain stages include footfall, zones, staff activity, and wait-time analysis; supporting branches cover advanced AI capabilities, dataset preparation and model training, and edge deployment. The README documents the workflow, but sampled backend excerpts omit much of the frame-processing implementation, so some internal links remain unverified.
Main entry points: main.py FastAPI backend · analytics.py Flow analytics · deployment.py Edge deployment · pro.html Web dashboard
Components 15
API and dashboard
- FastAPI backend
app/main.py - Web dashboard
app/pro.html
Video analysis
- Person detection
utils/detection.py - Object tracking
utils/tracking.py
Restaurant insights
- Flow analytics
utils/analytics.py - Advanced analytics
utils/advanced_analytics.py - Plots and heatmaps
utils/visualization.py
AI and training
- AI features
utils/ai_features.py - Extended AI
utils/more_ai.py - Dataset tools
utils/dataset.py - Model training
utils/train.py
Deployment and storage
- Edge deployment
utils/deployment.py - Analytics storage
utils/storage.py
Other components
- Restaurant operator
- Video feed
Connections 16
- Restaurant operator → FastAPI backend: requests analysis
README.md - Video feed → FastAPI backend: submits frames
README.md - FastAPI backend → Person detection: detects people
README.md - FastAPI backend → Object tracking: tracks people
README.md - FastAPI backend → Flow analytics: computes insights
README.md - FastAPI backend → Web dashboard: serves dashboard
app/main.py - Restaurant operator → Web dashboard: views results
README.md - Flow analytics → Plots and heatmaps: visualizes results
README.md - FastAPI backend → Advanced analytics: may use
- FastAPI backend → AI features: may use
- FastAPI backend → Extended AI: may use
- Dataset tools → Model training: prepares data
- Edge deployment → Person detection: integrates
utils/deployment.py - Edge deployment → Object tracking: integrates
utils/deployment.py - Edge deployment → Flow analytics: integrates
utils/deployment.py - Flow analytics → Analytics storage: persists metrics