AI Paper Atlas

Research map / Label-efficient and robust learning / Semi-supervised learning

Learning from Noisy Pseudo Labels for Semi-Supervised Temporal Action Localization

Kun Xia, Le Wang, Sanping Zhou, Gang Hua, Wei Tang

ICCV 2023 · Label-efficient and robust learning › Semi-supervised learning · approach: Pseudo-labeling for specialized tasks · 23 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 Noisy Label Ranking (confidence divided by boundary variance) (ablation)ActionFormer self-training · THUMOS14, 40% labeledaverage mAP (%): 47.1 -> 48.5Improved
add Noisy Label Filtering (adaptive class-balanced selection) (ablation)ActionFormer self-training · THUMOS14, 40% labeledaverage mAP (%): 48.5 -> 50.3Improved
add Noisy Label Learning (temporal consistency loss) (ablation)ActionFormer self-training · THUMOS14, 40% labeledaverage mAP (%): 50.3 -> 51.5Improved
add all three NPL components at 10% labels (ablation)ActionFormer self-training · THUMOS14, 10% labeledaverage mAP (%): 16.8 -> 20.3Improved
replace single threshold 0.3 with adaptive filtering (ablation)NPL · THUMOS14, 40% labeledaverage mAP (%): 49.2 -> 51.5Improved
replace fixed top-10 with adaptive filtering (ablation)NPL · THUMOS14, 40% labeledaverage mAP (%): 49.7 -> 51.5Improved
add temporal consistency loss (ablation)NPL · THUMOS14, 60% unlabeledaverage mAP (%): 44.7 -> 45.9Improved
vary Npos from 15 to 5 or 30 (ablation)NPL · THUMOS14, 40% labeledaverage mAP (%): 51.5 -> 51.0 (Npos 5), 50.7 (Npos 30)Hurt
apply NPL to pseudo labels at 10% labeledActionFormer · THUMOS14, 10% labeledaverage tIoU of pseudo labels (%): 27.4 -> 36.2Improved
add NPL to semi-supervised ActionFormer at 10% on ActivityNetActionFormer · ActivityNet v1.3 val, 10% labeledaverage mAP (%): 28.8 -> 32.5Improved
score pseudo labels by boundary variance instead of foreground scoreNoisy Label Ranking · THUMOS14 pseudo labelsPearson correlation with IoU: 0.449 -> 0.496Improved

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