Placeholder chapter page. Sections are produced by the book-skills pipeline.
Sections
- 20.1 From time-series FMs to raw multi-sensor FMs
- 20.2 Large sensor models: Google LSM and LSM-2 (adaptive/inherited masking, learning from incomplete data)
- 20.3 Apple-style wearable biosignal foundation models (PPG/ECG/accelerometer SSL at scale)
- 20.4 Motion/HAR foundation models
- 20.5 Handling missingness, multi-rate, and multi-device inputs
- 20.6 Personalization and adaptation of a pretrained sensor FM
- 20.7 Compute, privacy, and on-device constraints
Lab 20
fine-tune a pretrained sensor encoder for a new wearable task with limited labels; evaluate cross-user generalization.