AI Paper Atlas

Research map / Video / Action recognition

Towards Efficient General Feature Prediction in Masked Skeleton Modeling

Shengkai Sun, Zefan Zhang, Jianfeng Dong, Zhiyong Cheng, Xiaojun Chang, Meng Wang

ICCV 2025 · Video › Action recognition · approach: Skeleton-based action recognition · 5 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 motion-difference input for TGN instead of raw joints (ablation)GFP · NTU-60 x-viewtop-1 accuracy: 90.9 -> 92.0Improved
use unmasked motion input instead of masked joint input for TGN (ablation)GFP · NTU-60 x-viewtop-1 accuracy: 90.3 -> 92.0Improved
use 3-layer TGN instead of shallow 2-layer (ablation)GFP · NTU-60 x-viewtop-1 accuracy: 92.0 -> 91.8No clear effect
use 3-layer TGN instead of wider 4-layer (ablation)GFP · NTU-60 x-viewtop-1 accuracy: 92.0 -> 91.3No clear effect
use 3-layer projector instead of 2 layers (ablation)GFP · NTU-60 x-viewtop-1 accuracy: 91.3 -> 92.0Improved
use 3-layer projector instead of 4 layers (ablation)GFP · NTU-60 x-viewtop-1 accuracy: 91.6 -> 92.0Improved
add global and local hierarchical targets via TGN rather than removing them (ablation)GFP · NTU-60 x-subtop-1 accuracyImproved
add TGN-learned targets instead of direct low-level target prediction (ablation)GFP · NTU-60 x-subtop-1 accuracyImproved
replace SkeletonMAE-style reconstruction with hierarchical feature predictionGFP vs SkeletonMAE · NTU-60 pretrainingtraining speedup: 1x -> 6.2xImproved
replace patch-level EMA targets (S-JEPA) with lightweight TGNGFP vs S-JEPA · NTU-60 x-viewtop-1 accuracy: 89.8 -> 92.0Improved
replace MAMP motion targets with GFP high-level targetsGFP vs MAMP · NTU-60 retrieval x-viewtop-1 accuracy: 70.0 -> 87.1Improved
use GFP instead of S-JEPAGFP vs S-JEPA · NTU-120 x-subtop-1 accuracy: 79.6 -> 79.1Hurt

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