Research map / Label-efficient and robust learning / Recognition and long-tail
Adaptive Discriminative Regularization for Visual Classification
Qingsong Zhao, Yi Wang, Shuguang Dou, Chen Gong, Yin Wang, Cairong Zhao
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add ADR to cross entropy | ResNet-50 · ImageNet-1K | top-1 error: 23.06 -> 22.49 | Improved |
| add ADR to cross entropy | ResNet-101 · ImageNet-1K | top-1 error: 21.30 -> 20.76 | Improved |
| add ADR to label smoothing | ViT-B/16 · ImageNet-1K | top-1 error: 18.05 -> 17.69 | Improved |
| add ADR to CE, LS or OLS | ResNet-50 · Flowers-102 | top-1 accuracy: 90.89 -> 92.36 | Improved |
| add ADR to cross entropy | AlexNet · CIFAR-10 | top-1 accuracy: 73.33 -> 77.07 | Improved |
| add ADR to cross entropy | ConvMixer · CIFAR-10 | top-1 accuracy: 89.79 -> 92.47 | Improved |
| add ADR on top of label smoothing | AlexNet · CIFAR-10 | top-1 accuracy: 76.31 -> 77.75 | Improved |
| replace entropy regularizer with ADR (ablation) | AlexNet CE · CIFAR-10 | top-1 accuracy: 74.21 -> 77.07 | Improved |
| add entropy regularizer to CE (ablation) | AlexNet · CIFAR-10 | top-1 accuracy: 73.33 -> 74.21 | No clear effect |
| add ADR under 20% label noise (ablation) | CE · CIFAR-10 noisy labels | top-1 accuracy: 72.51 -> 74.89 | Improved |
| add ADR under 40% label noise (ablation) | CE · CIFAR-10 noisy labels | top-1 accuracy: 68.95 -> 70.53 | Improved |
| add ADR under 60% and 80% label noise (ablation) | CE · CIFAR-10 noisy labels | top-1 accuracy: 63.45 -> 63.65 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- Arcface — ArcFace: Additive Angular Margin Loss for Deep Face Recognition (CVPR 2019) · better on 1 of 1 shared result
- Cosface — CosFace: Large Margin Cosine Loss for Deep Face Recognition (CVPR 2018) · better on 1 of 1 shared result
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
- Rethinking Class-Balanced Methods for Long-Tailed Visual Recognition From a Domain Adaptation Perspective (CVPR 2020 · similarity 0.88)
- Confusion-Aware Spectral Regularizer for Long-Tailed Recognition (CVPR 2026 · similarity 0.87)
- I Am Going MAD: Maximum Discrepancy Competition for Comparing Classifiers Adaptively (ICLR 2020 · similarity 0.87)
- Unsupervised Learning for Class Distribution Mismatch (ICML 2025 · similarity 0.86)
- Generative Classifiers Avoid Shortcut Solutions (ICLR 2025 · similarity 0.86)
- Beyond Words: Augmenting Discriminative Richness via Diffusions in Unsupervised Prompt Learning (CVPR 2025 · similarity 0.86)
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.
