AI Paper Atlas

Research map / Video / Action recognition

View Adaptive Neural Networks for High Performance Skeleton-Based Human Action Recognition

Pengfei Zhang, Cuiling Lan, Junliang Xing, Wenjun Zeng, Jianru Xue, Nanning Zheng

TPAMI 2019 · Video › Action recognition · approach: Skeleton-based action recognition · 537 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 view adaptation subnetwork (rotation and translation) (ablation)S-trans + RNN · NTU RGB+D cross-viewaccuracy (%): 82.3 -> 87.6Improved
add view adaptation subnetwork (rotation and translation) (ablation)S-trans + RNN · NTU RGB+D cross-subjectaccuracy (%): 76.0 -> 79.4Improved
learn rotation only versus translation only (ablation)VA-RNN · NTU RGB+D cross-viewaccuracy (%): VA-trans 84.9, VA-rota 87.1, both 87.6Mixed
learned view adaptation instead of human-defined S-trans&S-rota (ablation)RNN · NTU RGB+D cross-viewaccuracy (%): 85.4 -> 87.6Improved
use frame-level instead of sequence-level pre-processing (ablation)RNN · NTU RGB+D cross-subjectaccuracy (%): 76.4 -> 74.1Hurt
add sequence-level translation to raw skeletons (ablation)RNN · NTU RGB+D cross-subjectaccuracy (%): 66.3 -> 76.0Improved
add random rotation data augmentation (ablation)VA-CNN · UWA3Daccuracy (%): 71.2 -> 79.3Improved
add random rotation data augmentation (ablation)VA-CNN · N-UCLAaccuracy (%): 81.7 -> 86.6Improved
add random rotation data augmentation (ablation)S-trans + CNN · NTU RGB+D cross-viewaccuracy (%): 92.2 -> 93.5Improved
add view adaptation on top of augmentation (ablation)S-trans + CNN (aug.) · UWA3Daccuracy (%): 77.0 -> 79.3Improved
stack more LSTM layers (4 to 6) in baseline (ablation)S-trans + RNN (aug.) · NTU RGB+D cross-viewaccuracy (%): 85.0 (3 layers) -> 83.9, 84.2, 84.4No clear effect
use deeper ResNet backbone (50 to 152) (ablation)S-trans + CNN (aug.) · NTU RGB+D cross-viewaccuracy (%): 93.5 -> 93.4No clear effect

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