[Badges]
AI-powered retail analytics platform that tracks visitor journeys, zone engagement, queue behavior, conversions, and revenue insights using Computer Vision, Spring Boot, PostgreSQL, and React.
- Overview
- Key Features
- System Architecture
- Tech Stack
- Database Design
- API Reference
- Computer Vision Pipeline
- Analytics & Metrics
- Project Structure
- Running Locally
- Docker Deployment
- Team
[Detailed project explanation]
- YOLO11 Person Detection
- ByteTrack Multi-Object Tracking
- Entry/Exit Detection
- Re-identification Support
- Multi-Camera Analytics
- Visitor Counting
- Live Occupancy
- Zone Dwell Time
- Heatmaps
- Queue Monitoring
- Conversion Tracking
- Revenue Attribution
- Average Basket Value
- Conversion Rate
- Queue Abandonment Rate
- Zone Performance Analysis
- Queue Spike
- Conversion Drop
- Dead Zone Detection
[ASCII Architecture Diagram]
| Layer | Technology |
|---|---|
| Backend | Spring Boot 3 |
| Database | PostgreSQL |
| Cache | Redis |
| CV Pipeline | Python |
| Detection | YOLO11 |
| Tracking | ByteTrack |
| Migration | Flyway |
| Containerization | Docker |
[event] [session] [visitor] [store] [zone] [transaction] [anomaly]
Explain relationships.
POST /events/ingest
GET /stores/{id}/metrics
GET /stores/{id}/heatmap
GET /sessions
etc.
Video ↓ YOLO11 ↓ ByteTrack ↓ Zone Mapping ↓ Event Generator ↓ Spring Boot API ↓ PostgreSQL ↓ Dashboard
- Unique Visitors
- Live Visitors
- Conversion Rate
- Revenue
- Avg Dwell Time
- Queue Depth
- Queue Abandonment
- Zone Heatmaps
Retail-Store/ ├── backend/ ├── pipeline/ ├── datasets/ ├── docker-compose.yml └── README.md
mvn spring-boot:run
python run_pipeline.py ...
docker-compose build docker-compose up
| Role | Responsibility |
|---|---|
| Backend | Spring Boot APIs |
| CV Pipeline | YOLO + ByteTrack |
MIT License