# RuView
RuView (WiFi DensePose) transforms ordinary WiFi signals into a contactless sensing system that detects human presence, tracks body pose, monitors vital signs, and identifies activities. No cameras or wearables required. It leverages Channel State Information (CSI) from low-cost ESP32 sensors to perceive people through walls and in complete darkness.
## Capabilities
- **Presence detection and people counting** across multiple individuals simultaneously
- **17-point body pose estimation** (COCO keypoints) without video, achieving 92.9% accuracy with camera-supervised training
- **Vital signs monitoring**: breathing rate (6-30 breaths/min) and heart rate (40-120 bpm)
- **Activity classification and fall detection**
- **Through-wall sensing** up to 5 meters depth
- **Self-learning**: adapts to new environments in under 30 seconds using spiking neural networks
## Architecture
Runs entirely on-device with a total BOM of ~$140 (ESP32-S3 + Cognitum Seed secure enclave). Built primarily in [[Rust]], with Python and JavaScript components. Uses multi-frequency mesh with channel hopping across 6 WiFi bands.
The system includes cryptographically attested measurement history with a witness chain for persistent storage, and supports Docker deployment (multi-arch: amd64 + arm64) alongside embedded firmware.
## Why It Matters
Privacy-centric alternative to camera-based monitoring. Useful for eldercare, occupancy sensing, smart home automation, and health monitoring without the surveillance trade-off.
## References
- GitHub: https://github.com/ruvnet/RuView
- License: MIT
## Related
- [[Rust]]