Gait foundation model in Nature Medicine predicts health from 3D skeletal motion
bravo_abad · x · 2026-11-09
Gabet et al. published a gait foundation model in Nature Medicine: a masked autoencoder pretrained on 3D skeletal motion from 3,328 adults (5 tasks, depth camera), with no health labels needed.
The key test: even after baselines knew each person's age, BMI, height and abdominal fat, the learned movement features still significantly improved predictions for 823 of 3,210 measured traits in male participants (vs. 188 in the comparison group) — evidence that movement carries health signals beyond age and body size.
More from Research
- NVIDIA's ScaleEvals (CoRL 2026) scales robot evals with reset-free autonomous play, near-zero humans — gharik · 2026-11-09
- Learning to read rewires your brain and makes you worse at recognizing faces — RichardMCNgo · 2026-11-09
- Blogger questions weight-space programming: just frontier models recalling old papers — builderjaydub · 2026-11-09
- Owain Evans proposes a practitioner guide to subliminal learning and phantom transfer — OwainEvans_UK · 2026-11-09
- Owain Evans proposes a research direction: data poisoning risks during RSI — OwainEvans_UK · 2026-11-09
- Owain Evans flags open research question: can reward hacking emerge without ever being rewarded — OwainEvans_UK · 2026-11-09