Research map / Video / Video segmentation and tracking
Contrast-Phys: Unsupervised Video-Based Remote Physiological Measurement via Spatiotemporal Contrast
Zhaodong Sun, Xiaobai Li
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 rPPG spatial similarity (2x2 spatial resolution instead of 1x1) (ablation) | Contrast-Phys · UBFC-rPPG HR | MAE (bpm): 3.14 -> 0.64 | Improved |
| raise ST-rPPG spatial resolution from 2x2 to 4x4 (ablation) | Contrast-Phys · UBFC-rPPG HR | MAE (bpm): 0.64 -> 0.55 | No clear effect |
| raise ST-rPPG spatial resolution from 2x2 to 8x8 (ablation) | Contrast-Phys · UBFC-rPPG HR | MAE (bpm): 0.64 -> 0.60 | No clear effect |
| shorten ST-rPPG time length from 10s to 5s (ablation) | Contrast-Phys · UBFC-rPPG HR | RMSE (bpm): 1.00 -> 1.36 | Hurt |
| lengthen ST-rPPG time length from 10s to 30s (ablation) | Contrast-Phys · UBFC-rPPG HR | MAE (bpm): 0.64 -> 1.97 | Hurt |
| inject periodic flashing noise into training videos | Gideon2021 · UBFC-rPPG HR | MAE (bpm): 1.85 -> 22.47 | Hurt |
| inject periodic flashing noise into training videos | Contrast-Phys · UBFC-rPPG HR | MAE (bpm): 0.64 -> 0.74 | No clear effect |
| use Contrast-Phys instead of Gideon2021 unsupervised baseline | Contrast-Phys · OBF HR | MAE (bpm): 2.83 -> 0.51 | Improved |
| use Contrast-Phys instead of Gideon2021 unsupervised baseline | Contrast-Phys · MMSE-HR cross-dataset | RMSE (bpm): 11.55 -> 7.34 | Improved |
| use Contrast-Phys single forward pass instead of Gideon2021 double pass | Contrast-Phys · Training speed | frames per second: 387.87 -> 802.45 | Improved |
| use Contrast-Phys instead of Gideon2021 | Contrast-Phys · UBFC-rPPG HRV (LF/HF) | RMSE: 0.691 -> 0.395 | Improved |
Later papers that compared with this method
3 later accepted papers report Contrast-Phys+ as a baseline in their results tables; 3 of them report a better number on most of the results they share with it.
- PHASE-Net: Physics-Grounded Harmonic Attention System for Efficient Remote Photoplethysmography Measurement (CVPR 2026 · better on 2 of 2 shared results)
- PhysDiff: Physiology-based Dynamicity Disentangled Diffusion Model for Remote Physiological Measurement (AAAI 2025 · better on 2 of 2 shared results)
- Non-Contrastive Unsupervised Learning of Physiological Signals From Video (CVPR 2023 · 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
- Contrast-Phys+: Unsupervised and Weakly-Supervised Video-Based Remote Physiological Measurement via Spatiotemporal Contrast (TPAMI 2024 · similarity 0.98)
- The Way to My Heart Is Through Contrastive Learning: Remote Photoplethysmography From Unlabelled Video (ICCV 2021 · similarity 0.95)
- Remote Photoplethysmography in Real-World and Extreme Lighting Scenarios (CVPR 2025 · similarity 0.91)
- PhysDiff: Physiology-based Dynamicity Disentangled Diffusion Model for Remote Physiological Measurement (AAAI 2025 · similarity 0.90)
- RhythmMamba: Fast, Lightweight, and Accurate Remote Physiological Measurement (AAAI 2025 · similarity 0.90)
- FacePhys: State of the Heart Learning (NeurIPS 2026 · similarity 0.89)
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.
