berkus
/
wifi-densepose
mirror of https://github.com/ruvnet/wifi-densepose.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
wifi-densepose/README.md

3.1 KiB
Raw Blame History

RuView — Unified WiFi Sensing Platform

RuView is a state-of-the-art platform for passive human sensing using WiFi CSI (Channel State Information). It transforms standard WiFi signals into high-resolution pose estimation, presence detection, and vital sign monitoring.

🚀 Quick Start (All-in-One CLI)

The ruview.py tool is your unified entry point for everything in the ecosystem.

# 1. Check if your hardware is connected
./ruview.py status

# 2. Start the sensing server (V2 Rust Engine)
./ruview.py sense server

# 3. View the live RF spectrum
./ruview.py scan live

# 4. Generate a Full ADR-031 System Report
./ruview.py bench report

📂 Project Organization

Folder Contents
v2/ The Core Engine. High-performance Rust implementation (ADR-117+). Includes the sensing server, pose estimation, and CLI.
models/ Pre-trained Hugging Face models (DensePose, CSI Encoders, Edge-optimized variants).
scripts/ Python and Node.js utilities for provisioning, data capture, and legacy watching.
ui/ Modern React-based desktop and web dashboard.
firmware/ ESP32-S3 and ESP32-C6 firmware source (Wi-Fi 6, iTWT, and mesh sync support).
docs/ Full architectural decision records (ADRs), release notes, and user guides.
aether-arena/ Official SOTA benchmarking and room calibration tools.

📡 Hardware: Flipper Zero WiFi Dev Board

Your dev board is currently active and streaming.

  • IP Address: 192.168.1.129
  • Target Port: 5005 (Production Sensing)
  • Status: Good and Going.

Optimization Commands

To get the most out of your hardware:

# Calibrate for your specific room (30s capture)
./ruview.py tune calibrate

# Setup HE (WiFi 6) Mesh (C6 Boards only)
./ruview.py node mesh

🛠️ V2 Support & Features

Yes, this project fully supports RuView V2 (ADR-117). The V2 engine is built in Rust for 100x lower latency and includes:

  • BFLD (Privacy Layer): Structural prevention of identity leakage.
  • Multistatic Fusion: High-accuracy 3D pose estimation from multiple nodes.
  • Apple Home / Matter: Native integration without needing Home Assistant.
  • 75K-param Transformer: Beats academic SOTA while running on a single CPU thread.

📖 Commands Reference

node

  • flash: Flash the latest firmware to your board.
  • provision: Set WiFi SSID/Pass and Target IP.
  • mesh: Enable high-resolution HE (WiFi 6) mesh mode.

scan

  • live: Real-time spectrum visualization.
  • wide: Wideband multi-frequency environment scanning.

sense

  • server: Start the V2 Rust Sensing Engine (Port 5005).
  • watcher: Simple Python-based presence alert watcher.

bench

  • report: Generate a full SOTA performance report.
  • aa: Run the AetherArena official score runner.

tune

  • calibrate: Record a room-baseline for improved accuracy.
  • adapt: Apply few-shot LoRA adaptation to your room.

Built by rUv. Released under MIT/Apache-2.0.