AI Paper Atlas

Research map / Video / Action recognition

Punching Bag vs. Punching Person: Motion Transferability in Videos

Raiyaan Abdullah, Jared Claypoole, Michael Cogswell, Ajay Divakaran, Yogesh Rawat

ICCV 2025 · Video › Action recognition · approach: Disentangled action recognition · 1 citations (OpenAlex)

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
replace realistic backgrounds with solid plain backgrounds in Syn-TA (ablation)EZ-CLIP · Syn-TA coarse motionsharmonic mean: 68.38 -> 77.17Improved
replace realistic backgrounds with solid plain backgrounds in Syn-TA (ablation)EZ-CLIP · Syn-TA fine motionsharmonic mean: 38.71 -> 51.01Improved
plain background for all 8 unimodal models (ablation)unimodal models · Syn-TA coarse motionsharmonic mean: 41.30 -> 51.62 (ResNet50)Improved
add coarse/fine disentanglement branches with residual fine-to-coarse linksEZ-CLIP · Syn-TA coarseHM: 67.96 -> 70.69 (Set 1), 68.80 -> 72.03 (Set 2)Improved
add coarse/fine disentanglement branchesEZ-CLIP · SSv2-TA coarseHM: 44.83 -> 46.89 (Set 1), 44.61 -> 47.82 (Set 2)Improved
add coarse/fine disentanglement branchesEZ-CLIP · K400-TA coarseHM: 76.07 -> 76.06 (Set 1), 74.80 -> 75.20 (Set 2)No clear effect
move from known to unknown fine contextsall 13 models · Syn-TA, K400-TA, SSv2-TA coarseaccuracy drop: AIM Syn-TA 99.13 -> 70.16Hurt
increase model sizeaction recognition models · SSv2-TAcoarse accuracyNo clear effect
increase model sizeaction recognition models · K400-TAcoarse accuracyImproved
use fine captions instead of coarse for unknown classesEZ-CLIP · K400-TA unknownaccuracy: <30% coarse -> >70% fineImproved
evaluate Syn-TA instead of K400-TA11 of 13 models · coarse harmonic meanharmonic meanHurt

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