AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Progressive Purification for Instance-Dependent Partial Label Learning

Ning Xu, Biao Liu, Jiaqi Lv, Congyu Qiao, Xin Geng

ICML 2023 · Label-efficient and robust learning › Recognition and long-tail · approach: Partial and multi-label learning · 5 citations (OpenAlex)

Read the paper

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 progressive purification to ProDEN (ablation)ProDEN · CIFAR-10accuracy: 76.77±0.63% -> 79.00±0.28%Improved
Add progressive purification to RCPLL (ablation)RCPLL · CIFAR-100accuracy: 55.18±0.70% -> 57.66±0.11%Improved
Add progressive purification to CCPLL (ablation)CCPLL · CIFAR-10accuracy: 72.87±0.82% -> 77.03±0.58%Improved
Add progressive purification to LWS (ablation)LWS · CIFAR-10accuracy: 46.77±0.66% -> 48.54±0.04%Improved
Add progressive purification to CAVL (ablation)CAVL · CIFAR-10accuracy: 62.69±1.65% -> 63.12±0.23%Improved
Add progressive purification to CLPL (ablation)CLPL · CIFAR-10accuracy: 53.61±0.31% -> 54.22±0.51%Improved
Add progressive purification to ProDEN (ablation)ProDEN · Lostaccuracy: 76.47±0.25% -> 78.57±0.45%Improved
Add progressive purification to RCPLL (ablation)RCPLL · MSRCv2accuracy: 49.47±0.43% -> 51.18±0.59%Improved
Add progressive purification to CCPLL (ablation)CCPLL · BirdSongaccuracy: 69.90±0.58% -> 71.50±0.06%Improved
Use progressive purification rather than the strongest listed baselinePartial-label learning methods · CIFAR-10accuracy: 86.11±0.43% -> 93.00±0.26%Improved
Use progressive purification rather than the strongest listed baselinePartial-label learning methods · CIFAR-100accuracy: 71.48±0.12% -> 71.82±0.08%Improved
Use progressive purification rather than the strongest listed baselinePartial-label learning methods · MSRCv2accuracy: 49.97±0.43% -> 45.86±0.28%Hurt

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