WiFi DensePose: How an Open-Source Tool Turns Ordinary Routers Into Through-Wall Body Trackers
An open-source project called WiFi DensePose — now rebranded and expanded as "RuView" — has gone viral on GitHub for demonstrating something that sounds like science fiction: reconstructing a full human body pose through a solid wall using nothing but ordinary WiFi signals. No camera, no LiDAR, no radar — just a router.
The Science: From Camera Pixels to Radio Waves
Human pose estimation — mapping the position of a person's limbs and body shape — has traditionally relied on cameras, LiDAR, or radar. Carnegie Mellon researchers Jiaqi Geng, Dong Huang, and Fernando De la Torre showed that this same dense body mapping could be achieved using only the radio signals already broadcast by commodity WiFi routers, even when a wall stands between the subject and the receiver.
Their method exploits Channel State Information (CSI) — metadata that WiFi hardware constantly generates to measure how a signal degrades between a transmitter and receiver. When a person moves through that signal's path, their body absorbs, reflects, and scatters the radio waves in measurable ways. The CMU team trained a neural network — architecturally similar to a U-Net, paired with a modified DensePose-RCNN model borrowed from computer vision — to translate those distorted signal patterns (phase and amplitude) into UV coordinates across 24 regions of the human body, effectively painting a 3D-style pose map from invisible radio noise.
The original lab setup used two ordinary TP-Link routers, each with three antennas — hardware costing roughly $30 a unit — proving the concept could work with equipment already sitting in millions of homes, without any specialised, expensive sensing hardware.
From Research Paper to Real-World Tool: Enter RuView
In late February 2026, a GitHub repository published by developer Reuven Cohen under the name RuView (built on the earlier WiFi-DensePose codebase) surged to the top of GitHub's trending list. Unlike the original academic prototype, RuView is described as a production-oriented, edge-AI implementation — meaning it is designed to actually run continuously in the field, not just inside a research lab.
| Feature | Reported Capability |
|---|---|
| Core hardware | Runs on an $8 ESP32 microcontroller; full model reportedly fits in 55 KB of memory |
| Signal capture rate | Captures CSI disturbances at up to 54,000 frames per second using a Rust-based pipeline |
| Output | Full-body pose tracking, presence detection, and claimed vital-sign (cardiac/respiratory) monitoring |
| Deployment | Distributed via Docker container, allowing rapid setup near any WiFi access point |
| Stated use case | Marketed for disaster response, elder care, and smart-building monitoring |
Why Security Researchers Are Alarmed
Cybersecurity outlets covering the project have been blunt about the dual-use risk. Multiple reports describe a simple attack scenario: a bad actor places a low-cost ESP32 node in a building's common area or near an existing WiFi access point, deploys the tool, and begins silently mapping the movement patterns, daily routines, and even biometric signals of occupants — entirely through walls, with no visual recording and minimal physical footprint to detect.
Security researchers have suggested initial mitigations: RF shielding for sensitive facilities, active monitoring for unauthorised ESP32-class devices on a network, and updated regulatory frameworks that explicitly extend surveillance law to cover CSI-based human tracking before the technology becomes widespread.
Not Entirely New — But Now Far More Accessible
This is not the industry's first brush with WiFi-based sensing. Some consumer ISPs already use simplified versions of this concept — for instance, motion-detection features in certain home WiFi systems that flag movement by analysing signal disruption between devices, without identifying individuals or producing images. Researchers also point to existing radar-based through-wall detection systems used by law enforcement and military units. What sets WiFi DensePose/RuView apart is the leap from coarse motion detection to dense, full-body pose reconstruction, built on hardware costing only a few dollars and released as open-source code anyone can download.
The Forensic and Legal Angle
For students and professionals tracking India's evolving surveillance and privacy jurisprudence, this development is relevant context. India's right to privacy was constitutionally affirmed by the Supreme Court in K.S. Puttaswamy v. Union of India, which held that any state infringement on privacy must satisfy tests of legality, necessity, and proportionality. Existing interception law in India, however, is built primarily around telecommunication content and metadata — not around passive, camera-free physical sensing technologies of this kind.
Tools like RuView raise a question current Indian surveillance and data-protection frameworks were not designed to answer: how should the law treat a device that produces no photograph, no video, and no traditional "recording," yet still reconstructs the precise physical movement and biometric signals of a person inside their own home? Legal commentators tracking India's surveillance regime have already flagged that existing safeguards struggle even with conventional interception, let alone with emerging RF-based sensing that falls outside current statutory definitions.
What to Watch Next
- Whether commercial router and ISP manufacturers begin shipping similar dense-sensing features by default.
- Regulatory responses from cybersecurity and data-protection authorities regarding CSI-based tracking.
- Development of detection tools capable of identifying rogue sensing nodes on a network.
- Academic and forensic debate over whether RF-derived pose data could ever be treated as admissible evidence, and under what safeguards.
This report is compiled from publicly available research papers, technical news coverage, and the project's public GitHub repository for educational and awareness purposes. Budding Forensic Expert does not endorse or facilitate the deployment of covert surveillance technology.
Sources
- CMU Research Paper — "DensePose From WiFi" (arXiv): arxiv.org/abs/2301.00250
- Full Paper PDF: arxiv.org/pdf/2301.00250
- RuView GitHub Repository: github.com/ruvnet/wifi-densepose
- Cybersecurity News — "WiFi Signals Reveal Human Activities Through Walls": cybersecuritynews.com
- GBHackers — "WiFi Signals Can Track Human Activity Through Walls": gbhackers.com
- Cybernews — "Viral GitHub project claims WiFi can 'see through walls'": cybernews.com
- Kaspersky Official Blog — "Human body pose recognition using Wi-Fi signal": kaspersky.com/blog
- Medium (Ajay Verma) — "WiFi Can See You Now: The Privacy Paradox": medium.com
- Internet Governance Project — "Privacy in Peril: India's Interception Regime": internetgovernance.org
- ORF — "The State of Surveillance in India": orfonline.org
