AI Paper Atlas

Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning

Reading Your Actions: Learning Generalizable Action Representations via Pre-training AEMG

Zhenghao Huang, Huilin Yao, Kaikai Wang, Lin Shu

CVPR 2026 · Label-efficient and robust learning › Zero-/few-shot and meta-learning · approach: Zero-shot robustness and transfer

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 EMG vocabulary with vector quantization (ablation)AEMG-Large · ULB-MLG LOSOaccuracy (%): 62.3 -> 91.50Improved
add masked modeling pre-training (ablation)AEMG-Large · ULB-MLG LOSOaccuracy (%): 81.3 -> 91.50Improved
add masked modeling pre-training (ablation)AEMG-Large · Ninapro DB4 LOSOaccuracy (%): 79.5 -> 88.10Improved
use NCT contraction-based tokenization instead of fixed windows (ablation)AEMG · 4 datasets LOSO averageaccuracy (%): 82.15 -> 91.28Improved
use NCT instead of fixed windows (ablation)AEMG · VQ reconstructionreconstruction lossImproved
use NCT instead of fixed windows (ablation)AEMG · masked pre-trainingmask prediction accuracyHurt
scale from AEMG-Base to AEMG-LargeAEMG · 4 datasets LOSO averageaccuracy (%): 86.49 -> 89.81Improved
fine-tune with only 5% target data instead of 80%AEMG-Large · ULB-MLGaccuracy (%): 91.00 -> 88.50Mixed
use AEMG-Large instead of best domain adaptation baseline on EMG-EPNAEMG-Large · EMG-EPN LOSOaccuracy (%): 88.97 -> 88.32Hurt

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