wifi-densepose

Commit Graph

Author	SHA1	Message	Date
ruv	3b4e151507	docs: ADR-081 add ruvector-cnn spectrogram gesture classification - Replace DTW with CNN on CSI spectrograms via ruvector-cnn WASM - Pipeline: CSI → STFT → 64x64 spectrogram → CnnEmbedder → 128-dim → classifier - Two-phase training: InfoNCE contrastive + supervised classification - Dual-path fusion: DTW + CNN in parallel for max robustness - Comparison table: CNN ~95% vs DTW ~85% accuracy (literature) - Fallback: lightweight 1D CNN for ESP32 edge deployment Co-Authored-By: claude-flow <ruv@ruv.net>	2026-04-07 09:04:16 -04:00
ruv	68d47a25d5	docs: ADR-081 add AR camera overlay with floating charts + lower third - AR overlay: live camera feed with skeleton, gesture cursor, and floating charts anchored to hand/body position - Lower third: RuView "pi" logo, vital signs, gesture state, sensor status in broadcast-style bar (semi-transparent dark, teal accents) - 6 composited layers: camera → skeleton → cursor → chart → labels → lower third - Chart placement rules: follows dominant hand, stays in frame bounds - Skeleton style: teal keypoints/bones, yellow highlight on active hand - Cursor types: open hand, pointing ray, grab, pinch, ghost (CSI-only) Co-Authored-By: claude-flow <ruv@ruv.net>	2026-04-07 08:57:41 -04:00
ruv	0d3292314b	docs: ADR-081 gesture-controlled data visualization Camera + CSI fusion for hands-free chart manipulation: - 11 arm-level gestures (CSI-detectable): swipe, circle, hold, spread - 7 finger-level gestures (camera-required): pinch, point, grab, thumbs - Fusion engine: camera precision + CSI through-wall capability - Chart types: line, bar, 3D scatter, heatmap, gauge, spectrogram - Visual feedback: gesture cursor overlay + state indicator - WebSocket protocol for gesture events → UI commands - Dual-mode: fusion (full precision) or CSI-only (works in dark) - Builds on WiFlow (ADR-079) + DTW gestures (ADR-029) Co-Authored-By: claude-flow <ruv@ruv.net>	2026-04-07 08:52:39 -04:00

Author

SHA1

Message

Date

ruv

3b4e151507

docs: ADR-081 add ruvector-cnn spectrogram gesture classification

- Replace DTW with CNN on CSI spectrograms via ruvector-cnn WASM
- Pipeline: CSI → STFT → 64x64 spectrogram → CnnEmbedder → 128-dim → classifier
- Two-phase training: InfoNCE contrastive + supervised classification
- Dual-path fusion: DTW + CNN in parallel for max robustness
- Comparison table: CNN ~95% vs DTW ~85% accuracy (literature)
- Fallback: lightweight 1D CNN for ESP32 edge deployment

Co-Authored-By: claude-flow <ruv@ruv.net>

2026-04-07 09:04:16 -04:00

ruv

68d47a25d5

docs: ADR-081 add AR camera overlay with floating charts + lower third

- AR overlay: live camera feed with skeleton, gesture cursor, and
  floating charts anchored to hand/body position
- Lower third: RuView "pi" logo, vital signs, gesture state, sensor
  status in broadcast-style bar (semi-transparent dark, teal accents)
- 6 composited layers: camera → skeleton → cursor → chart → labels → lower third
- Chart placement rules: follows dominant hand, stays in frame bounds
- Skeleton style: teal keypoints/bones, yellow highlight on active hand
- Cursor types: open hand, pointing ray, grab, pinch, ghost (CSI-only)

Co-Authored-By: claude-flow <ruv@ruv.net>

2026-04-07 08:57:41 -04:00

ruv

0d3292314b

docs: ADR-081 gesture-controlled data visualization

Camera + CSI fusion for hands-free chart manipulation:
- 11 arm-level gestures (CSI-detectable): swipe, circle, hold, spread
- 7 finger-level gestures (camera-required): pinch, point, grab, thumbs
- Fusion engine: camera precision + CSI through-wall capability
- Chart types: line, bar, 3D scatter, heatmap, gauge, spectrogram
- Visual feedback: gesture cursor overlay + state indicator
- WebSocket protocol for gesture events → UI commands
- Dual-mode: fusion (full precision) or CSI-only (works in dark)
- Builds on WiFlow (ADR-079) + DTW gestures (ADR-029)

Co-Authored-By: claude-flow <ruv@ruv.net>

2026-04-07 08:52:39 -04:00

3 Commits