AI Paper Atlas

Research map / Video / Action recognition

SUGAR: Learning Skeleton Representation with Visual-Motion Knowledge for Action Recognition

Qilang Ye, Yu Zhou, Lian He, Jie Zhang, Xuanming Guo, Jiayu Zhang, Mingkui Tan, Weicheng Xie, Yue Sun, Tao Tan, Xiaochen Yuan, Ghada Khoriba and 1 more

AAAI 2026 · 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 and visual language knowledge to supervise the skeleton encoder (ablation)SUGAR · Toyota Smarthome overallaccuracy: 69.2 -> 73.4Improved
add visual knowledge alone (without motion) (ablation)SUGAR · Toyota Smarthome overallaccuracy: 69.2 -> 69.4No clear effect
add motion knowledge alone (ablation)SUGAR · Toyota Smarthome overallaccuracy: 69.2 -> 72.1Improved
add visual knowledge on top of motion knowledge (ablation)SUGAR · Toyota Smarthome overallaccuracy: 72.1 -> 73.4Improved
use TQP instead of a single Q-Former (ablation)SUGAR · Toyota Smarthome overallaccuracy: 70.7 -> 73.4Improved
use TQP instead of one linear layer (ablation)SUGAR · Toyota Smarthome overallaccuracy: 70.4 -> 73.4Improved
use TQP instead of cross-attention bridging (ablation)SUGAR · Toyota Smarthome overallaccuracy: 52.1 -> 73.4Improved
shorten action token sequence from full length 1000 down to moderate lengths (ablation)SUGAR · NTU RGB+D 60accuracyImproved
compress action tokens all the way to 1 (pooling) (ablation)SUGAR · NTU RGB+D 60accuracyHurt
use SUGAR (LLM recognizer with language-supervised skeleton) instead of LLM-ARLLM-based recognizer · Toyota Smarthome X-view1mean per-class accuracy: 36.1 -> 50.9Improved
use SUGAR instead of LLM-AR on NTU 60 X-view and NTU 120 X-viewLLM-based recognizer · NTU RGB+D X-viewaccuracy: NTU60 X-view 98.4 -> 97.8; NTU120 X-view 91.5 -> 89.7Mixed
use LLM recognizer (SUGAR) instead of linear classifier for unseen classesskeleton recognizer · Zero-shot unseen NTU120 classes, Protocol 1Top-1 accuracy: 34.5 -> 65.3Improved

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