Research map / Vision-language and multimodal / Cross-modal learning
RPM-Distill: Physiology-guided Adaptive Cross-modal Distillation for Robust Remote Physiological Measurement
Jiyao Wang, Qingyong Hu, Duoxun Tang, Xiao Yang, Kaishun Wu, Jiangbo Yu
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 L_peak (ablation) | RPM-Distill · EquiPleth (cross-dataset) | MAE (bpm): 4.30 -> 1.57 | Improved |
| add L_peak (ablation) | RPM-Distill · PhysDrive (cross-dataset) | MAE (bpm): 16.28 -> 14.15 | Improved |
| add L_off (ablation) | RPM-Distill · EquiPleth (cross-dataset) | MAE (bpm): 3.40 -> 1.57 | Improved |
| add L_shape (ablation) | RPM-Distill · EquiPleth (cross-dataset) | MAE (bpm): 3.13 -> 1.57 | Improved |
| add learned component weights (vs fixed equal weights) (ablation) | RPM-Distill · EquiPleth (cross-dataset) | MAE (bpm): 3.25 -> 1.57 | Improved |
| add output gate (ablation) | RPM-Distill · EquiPleth (cross-dataset) | MAE (bpm): 3.45 -> 1.57 | Improved |
| add spectral policy network and bilevel optimization (ablation) | RPM-Distill · EquiPleth (cross-dataset) | MAE (bpm): 2.78 -> 1.57 | Improved |
| use bilevel meta-optimization (vs joint training of policy) (ablation) | RPM-Distill · PhysDrive (cross-dataset) | MAE (bpm): 16.15 -> 14.15 | Improved |
| add gate under 100% training RF misalignment (ablation) | RPM-Distill · EquiPleth->PhysDrive, shift prob 1.0 | MAE (bpm): 18.08 -> 14.80 | Improved |
| increase labeled fraction from 40% to 100% | RPM-Distill · PhysDrive Night | MAE (bpm): 17.87 -> 18.02 | No clear effect |
| increase labeled fraction from 40% to 100% | RPM-Distill · PhysDrive Talking | MAE (bpm): 14.36 -> 15.99 | Hurt |
| add RPM-Distill with 40% labels vs FactorizePhys full labels | FactorizePhys · PhysDrive Noon | MAE (bpm): 13.28 -> 11.93 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- C2KD — C2KD: Bridging the Modality Gap for Cross-Modal Knowledge Distillation (CVPR 2024) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Spatial Alignment and Temporal Matching Adapter for Video-Radar Remote Physiological Measurement (ICCV 2025 · similarity 0.90)
- FreqPhys: Repurposing Implicit Physiological Frequency Prior for Robust Remote Photoplethysmography (ECCV 2026 · similarity 0.89)
- PhysDiff: Physiology-based Dynamicity Disentangled Diffusion Model for Remote Physiological Measurement (AAAI 2025 · similarity 0.87)
- Fusion-Vital: Video-RF Fusion Transformer for Advanced Remote Physiological Measurement (AAAI 2024 · similarity 0.87)
- RhythmMamba: Fast, Lightweight, and Accurate Remote Physiological Measurement (AAAI 2025 · similarity 0.87)
- 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.
