AI Paper Atlas

Research map / Video / Video segmentation and tracking

Continuous Heart Rate Variability Estimation from Egocentric Systems for Skill Assessment

Berken Utku Demirel, Christian Holz

ECCV 2026 · Video › Video segmentation and tracking · approach: Physics-inspired motion prediction

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 changedWhereMeasuredEffect
add low-high decomposition (ablation)3D backbone · egoPPG-DB HR (4 s)MAE (bpm): 21.76 -> 14.80Improved
add cross-domain pretraining (ablation)3D backbone + low-high · egoPPG-DB HR (4 s)MAE (bpm): 14.80 -> 12.72Improved
add cross-domain pretraining (ablation)3D backbone + low-high · egoPPG-DB IBI (4 s)MAE (ms): 117.72 -> 105.91Improved
use Fourier magnitude input in pretraining (ablation)EgoHRV · egoPPG-DB HR (4 s)MAE (bpm): 14.87 -> 12.72Improved
add NLL objective in pretraining (ablation)EgoHRV · egoPPG-DB HR (4 s)MAE (bpm): 15.34 -> 12.72Improved
add NLL objective in pretraining (ablation)EgoHRV · egoPPG-DB HR (4 s)Pearson r: 0.08 -> 0.42Improved
use U-Net instead of 3-layer MLP encoder (ablation)EgoHRV · egoPPG-DB HR (4 s)MAE (bpm): 13.40 -> 12.72Improved
replace PulseFormer with EgoHRVEgoHRV · egoPPG-DB HR (4 s)MAE (bpm): 22.10 -> 12.72Improved
replace PulseFormer with EgoHRVEgoHRV · egoPPG-DB SDNNMAE (ms): 33.71 -> 10.89Improved
add HRV features instead of 60 s mean HRTimeSformer proficiency · EgoExo4D (Ego)top-1 accuracy (%): 44.30 -> 46.75Improved
add uncertainty log-variance to HRV features (ablation)PhysFusion · EgoExo4D (Ego)top-1 accuracy (%): 45.31 -> 46.75Improved
add IMU input to uncertainty head (ablation)EgoHRV · uncertainty calibrationexpected calibration error (%): 20.0 -> 9.7Improved

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.

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

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy