AI Paper Atlas

Research map / Label-efficient and robust learning / Noisy labels

Self-Adaptive Training: beyond Empirical Risk Minimization

Lang Huang, Chao Zhang, Hongyang Zhang

NeurIPS 2020 · Label-efficient and robust learning › Noisy labels · approach: Loss correction and noise robustness theory · 87 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
Use self-adaptive training instead of ERMResNet-50 · ImageNet validation, 40% label noiseTop1 accuracy: 69.5 -> 71.5Improved
Use self-adaptive training instead of ERMCIFAR10 · CIFAR10-C corruption levels 1–5average accuracy: 88.44 -> 89.41; 83.22 -> 84.53; 77.26 -> 78.83; 70.40 -> 71.90; 58.91 -> 60.77Improved
Replace TRADES cross-entropy term with self-adaptive trainingTRADES · CIFAR10 test under white-box PGD-20robust accuracy: 1%~3%Improved
Remove sample re-weighting (ablation)Self-Adaptive Training · CIFAR10 and CIFAR100, 40% and 80% label noiseaccuracy: 92.64 -> 92.49; 78.58 -> 78.10; 71.38 -> 69.52; 38.72 -> 36.78Hurt
Remove exponential moving average (ablation)Self-Adaptive Training · CIFAR10 and CIFAR100, 40% and 80% label noiseaccuracy: 92.64 -> 72.00; 78.58 -> 28.17; 71.38 -> 50.93; 38.72 -> 11.57Hurt
Set α to 0.6 instead of 0.9 (ablation)Self-Adaptive Training · CIFAR10, 40% label noiseaccuracy: 92.64 -> 90.17Hurt
Set α to 0.8 instead of 0.9 (ablation)Self-Adaptive Training · CIFAR10, 40% label noiseaccuracy: 92.64 -> 91.91Hurt
Set α to 0.95 instead of 0.9 (ablation)Self-Adaptive Training · CIFAR10, 40% label noiseaccuracy: 92.64 -> 92.54No clear effect
Set α to 0.99 instead of 0.9 (ablation)Self-Adaptive Training · CIFAR10, 40% label noiseaccuracy: 92.64 -> 84.38Hurt
Set E_s to 20 instead of 60 (ablation)Self-Adaptive Training · CIFAR10, 40% label noiseaccuracy: 92.64 -> 89.58Hurt
Set E_s to 40 instead of 60 (ablation)Self-Adaptive Training · CIFAR10, 40% label noiseaccuracy: 92.64 -> 91.89Hurt
Set E_s to 80 instead of 60 (ablation)Self-Adaptive Training · CIFAR10, 40% label noiseaccuracy: 92.64 -> 92.26No clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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