Sense the
world through WiFi

Contactless human pose and vital sign detection.
No cameras. No wearables. No cloud.

How it works

WiFi signals already fill every room. Radia reads them.

01
WiFi Capture

ESP32 mesh nodes capture Channel State Information — 56 subcarriers at 28 Hz across channels 1, 6 & 11.

02
Signal Processing

SpotFi, Hampel & Fresnel algorithms strip room noise and isolate human motion in under 100μs per frame.

03
Radia AI Engine

Attention networks and graph algorithms map signal disturbances to 17 body keypoints and vital signs.

04
Live Output

Real-time pose, breathing (6–30 BPM) and heart rate (40–120 BPM) — through walls, no cameras, < 1ms latency.

terminal
$docker pull ruvnet/wifi-densepose:latest
$docker run -p 3000:3000 ruvnet/wifi-densepose:latest
$# Open http://localhost:3000
Note: This demo does not reflect the latest release of Radia. It is a demonstration of WiFi CSI sensing technology as a whole. All credit goes to the researchers and contributors who made this work possible.

Comparison

Radia vs the field

Feature
● Radia
ReVue
RF-Pose (MIT)
WiPose
Open source
✓
✓
✗
✗
Through-wall detection
✓
Partial
✓
✗
Public dataset
✓
✗
✗
✗
Vitals (heart + breath)
✓
✗
✗
✗
Docker deploy
✓
✓
✗
✗
Body keypoints
17
N/A
25
~14
No cloud required
✓
✓
✗
✗

Competitor data sourced from published research papers and public repositories. Verify before citing.

GitHubDocumentation →