AI Paper Atlas

Research map / Video / Action recognition

Beyond Label Semantics: Language-Guided Action Anatomy for Few-shot Action Recognition

Zefeng Qian, Xincheng Yao, Yifei Huang, Chongyang Zhang, Jiangyong Ying, Hong Sun

ICCV 2025 · Video › Action recognition · approach: Disentangled action recognition

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 visual anatomy (atomic action segmentation) (ablation)CLIP-FSAR · HMDB51 1-shotaccuracy: 75.8 -> 79.9Improved
add textual anatomy (atomic descriptions) (ablation)CLIP-FSAR · HMDB51 1-shotaccuracy: 75.8 -> 79.6Improved
add visual anatomy alone (ablation)CLIP-FSAR · HMDB51 5-shotaccuracy: 87.7 -> 86.0Hurt
add both visual and textual anatomy (ablation)LGA · HMDB51 1-shotaccuracy: 75.8 -> 80.8Improved
add video-text matching (ablation)LGA · HMDB51 1-shotaccuracy: 80.8 -> 86.8Improved
add video-text matching (ablation)LGA · HMDB51 5-shotaccuracy: 88.2 -> 89.3Improved
use atomic descriptions instead of action label with self-attention global fusion (ablation)fusion module · HMDB51 1-shotaccuracy: 79.3 -> 85.2Improved
use cross-attention instead of self-attention fusion (ablation)fusion module · HMDB51 1-shotaccuracy: 79.3 -> 81.7Improved
use local instead of global fusion with atomic descriptions (ablation)fusion module · SSv2-Small 5-shotaccuracy: 65.4 -> 66.5Mixed
use AB-MHM instead of Bi-MHM for video-video matching (ablation)LGA · SSv2-Small 1-shotaccuracy: 55.2 -> 58.9Improved
use CLUSTER-Segment instead of TW-FINCH (ablation)Visual Anatomy · HMDB51 1-shotaccuracy: 82.9 -> 86.8Improved
use CLUSTER-Segment instead of uniform HARD split (ablation)Visual Anatomy · SSv2-Small 1-shotaccuracy: 57.0 -> 58.9Improved

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