An open-source project called RuView is drawing intense attention and equally intense skepticism for its claim that a common Wi-Fi router paired with a roughly $5–9 chip can detect, track and estimate the posture of people through walls—without any camera.
Open-Source AI Hardware · Wi-Fi Sensing
Seeing Through Walls With a $9 Chip — Real Science, or Viral Hype?
RuView claims a common Wi-Fi router plus a cheap ESP32 board can detect, track and pose people through walls — no camera. The underlying Wi-Fi sensing is genuine research; its sweeping headline claims remain independently unverified.
$5–9
Cost of the ESP32-S3 chip paired with a commodity router
17
Joint points claimed for dense body-pose estimation
8KB
Size of the 4-bit quantized model version
17
Sensing applications listed, from breathing to fall detection
The Reported Benchmark
Held-out temporal-triplet accuracy vs. a perfect score — and the claim that was quietly withdrawn.
82.3%
REPORTED
Held-out triplet accuracy
100%
CLAIM WITHDRAWN
Earlier "100% presence" claim, now removed
Trained over 12.2M steps on 60,000 frames & 610,000 contrastive triplets — running at 164,183 embeddings/sec on an M4 Pro.
Supporters say
The core science of Wi-Fi CSI sensing is real and established
Camera-free, privacy-conscious occupancy detection is a genuinely useful direction
Basic presence and motion tasks can plausibly work in controlled labs
Critics say
Reliance on mock or synthetic CSI data and incomplete model implementation
Doubts over the authenticity of rapid GitHub star growth
An audit fork labels the setup a "non-functional facade"
The recurring tension
A legitimate research foundation gets wrapped in consumer-facing claims that outrun what has been independently verified.
Through-wall pose plus heart-rate tracking appears limited to tightly controlled setups; on ordinary routers reliability is low and it is not plug-and-play. Its future hinges on external replication of the benchmarks and proof of dependable operation outside curated conditions.
Continue reading The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in ✓ Signed in — this article isn’t included in your current plan.Unlocking the full article…