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

docs(study): cross-dataset transfer tested (MM-Fi -> NTU-Fi, honest negative) 

Tested the cross-dataset frontier: MM-Fi-trained CSI representation does NOT
transfer beneficially to NTU-Fi HAR (frozen probe 91.5% = random features
93%; full fine-tune 75% < probe). CSI reps are distribution-locked, same
root cause as within-MM-Fi cross-subject/-env collapse. Caveat: NTU-Fi 6
coarse activities are an easy target (random->93%). Updates the study's
cross-dataset limitation from 'untested' to this measured result.

Co-Authored-By: claude-flow <ruv@ruv.net>
This commit is contained in:
ruv 2026-05-31 03:27:38 -04:00
parent 695b5fb700
commit af68bd68d8
1 changed files with 9 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
docs/benchmarks/mmfi-wifi-sensing-study.md
View File

@ -110,9 +110,15 @@ architecture-agnostic LoRA on the pose head, tested).
## 5. Honest limitations ## 5. Honest limitations
- All generalization numbers are within MM-Fi (one dataset, one hardware setup). **Cross-*dataset*** - Most generalization numbers are within MM-Fi (one dataset, one hardware setup). **Cross-*dataset***
transfer (different radios/rooms/protocols) is untested — the next real frontier, pending a second transfer was tested against **NTU-Fi HAR** (same 3×114 layout, different lab/hardware/rooms): an
public dataset. MM-Fi-trained representation does **not** transfer beneficially — a frozen MM-Fi trunk probes NTU-Fi
at 91.5%, *no better than random features* (93%), and full fine-tuning (75%) underperforms a linear
probe. CSI representations are **distribution-locked** (same root cause as the within-MM-Fi
cross-subject/-environment collapse); the practical answer is on-target training/few-shot, not
transferable zero-shot features. Caveat: NTU-Fi's 6 coarse activities are an *easy* target (random
features → 93%), so it weakly stresses representation quality. A harder cross-dataset pose benchmark
remains open.
- Random-split numbers are reported only to compare to prior work on the same protocol; they are - Random-split numbers are reported only to compare to prior work on the same protocol; they are
in-domain and partly leaky. The cross-subject / cross-environment numbers are the honest ones. in-domain and partly leaky. The cross-subject / cross-environment numbers are the honest ones.
- Action-recognition accuracy is window-level (MM-Fi's own HAR experiment is clip-level); not directly - Action-recognition accuracy is window-level (MM-Fi's own HAR experiment is clip-level); not directly