AI Paper Atlas

Research map / Video / Action recognition

Joint Self-Supervised Video Alignment and Action Segmentation

Ali Shah Ali, Syed Ahmed Mahmood, Mubin Saeed, Andrey Konin, M. Zeeshan Zia, Quoc-Huy Tran

ICCV 2025 · Video › Action recognition · approach: Temporal action segmentation · 11 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
add structural prior (GW term) (ablation)VAOT · IKEA ASMAcc@1.0: 37.81 -> 38.64Improved
add temporal prior (ablation)VAOT · IKEA ASMAcc@1.0: 17.84 -> 38.64Improved
use balanced instead of full unbalanced assignment (ablation)VAOT · IKEA ASMAcc@1.0: 25.24 -> 38.64Improved
add virtual frame for background frames (ablation)VAOT · IKEA ASMAcc@1.0: 36.10 -> 38.64Improved
add structural prior (GW term) (ablation)VAOT · PouringKendall tau: 72.49 -> 88.28Improved
add temporal prior (ablation)VAOT · PouringKendall tau: 78.11 -> 88.28Improved
use balanced instead of full unbalanced assignment (ablation)VAOT · PouringKendall tau: 68.78 -> 88.28Improved
add virtual frame for background frames (ablation)VAOT · PouringKendall tau: 82.65 -> 88.28Improved
use VAOT (FGW with structural prior) instead of VAVA (Kantorovich with optimality prior)video alignment · IKEA ASMAcc@1.0: 29.10 -> 38.64Improved
add action segmentation jointly (VASOT instead of VAOT)video alignment · IKEA ASMAcc@1.0: 38.64 -> 31.02Hurt
add video alignment jointly (VASOT instead of ASOT)action segmentation · 50 Salads EvalMoF: 59.3 -> 60.6Improved
add video alignment jointly (VASOT instead of ASOT)action segmentation · Desktop AssemblyF1: 68.0 -> 75.1Improved

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