Research map / Label-efficient and robust learning / Domain adaptation and generalization
Align the GAP: Prior-Based Unified Multi-task Remote Physiological Measurement Framework For Domain Generalization and Personalization
Jiyao Wang, Xiao Yang, Hao Lu, Dengbo He, Kaishun Wu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add test-time personalized adaptation (GAP-P) on top of GAP-G | GAP-G · PURE HR | MAE: 5.27 -> 4.51 | Improved |
| add test-time personalized adaptation (GAP-P) on top of GAP-G | GAP-G · HMPC-D night HR | RMSE: 9.90 -> 7.27 | Improved |
| add prior-based augmentation and self-supervision (GAP-G) versus PhysMLE | PhysMLE · VIPL-HR SpO2 | MAE: 10.94 -> 5.06 | Improved |
| use GAP-G instead of PhysMLE for HR | PhysMLE · VIPL-HR HR | MAE: 7.53 -> 8.16 | Hurt |
| use GAP-G instead of PhysMLE for HR | PhysMLE · HCW HR | MAE: 7.13 -> 8.07 | Hurt |
| use GAP-G instead of PhysMLE (smaller, simpler model) | PhysMLE · Inference runtime | runtime (ms): 1780 -> 56 | Improved |
| apply generic TTA method TENT on top of GAP-G | GAP-G · public datasets | accuracy | Hurt |
| add individual modeling losses L_PE and L_P (ablation) | GAP-G · HMPC-D | MAE | Improved |
| add time-domain consistency loss L_TC (ablation) | GAP-G · HMPC-D | MAE | Improved |
| add frequency-domain consistency loss L_FC (ablation) | GAP-P · HMPC-D (TTPA) | MAE | Improved |
| add individual bias loss L_P and time-domain inconsistency loss L_TIC (ablation) | GAP-P · HMPC-D (TTPA) | MAE | Improved |
| raise orthogonality threshold pi from 0.0 to 0.3 (ablation) | GAP-G · HMPC-D HR and RR | MAE | Improved |
Closest papers by meaning
- PhysMLE: Generalizable and Priors-Inclusive Multi-Task Remote Physiological Measurement (TPAMI 2025 · similarity 0.93)
- Video-based Remote Physiological Measurement via Cross-verified Feature Disentangling (ECCV 2020 · similarity 0.90)
- RhythmMamba: Fast, Lightweight, and Accurate Remote Physiological Measurement (AAAI 2025 · similarity 0.87)
- Remote Photoplethysmography in Real-World and Extreme Lighting Scenarios (CVPR 2025 · similarity 0.86)
- PhysDiff: Physiology-based Dynamicity Disentangled Diffusion Model for Remote Physiological Measurement (AAAI 2025 · similarity 0.85)
- FacePhys: State of the Heart Learning (NeurIPS 2026 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
