Research map / Label-efficient and robust learning / OOD and anomaly detection
Boosting Out-of-distribution Detection with Typical Features
Yao Zhu, YueFeng Chen, Chuanlong Xie, Xiaodan Li, Rong Zhang, Hui Xue', Xiang Tian, bolun zheng, Yaowu Chen
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 |
|---|---|---|---|
| Apply BATS feature rectification with Energy scoring | ResNet-50 · ImageNet-1k OOD average | FPR95: 58.16 -> 27.11 | Improved |
| Apply BATS feature rectification with Energy scoring | ResNet-50 · ImageNet-1k OOD average | AUROC: 85.82 -> 94.28 | Improved |
| Apply BATS feature rectification with Energy scoring | ResNet-18 · CIFAR-10 OOD average | FPR95: 32.99 -> 29.96 | Improved |
| Apply BATS feature rectification with Energy scoring | ResNet-18 · CIFAR-100 OOD average | FPR95: 58.12 -> 52.73 | Improved |
| Apply BATS feature rectification | Energy score on ImageNet-1k, ResNet-50 · ImageNet-1k OOD average | FPR95: 58.16 -> 27.11 | Improved |
| Apply BATS feature rectification | GradNorm on ImageNet-1k, ResNet-50 · ImageNet-1k OOD average | FPR95: 40.08 -> 21.53 | Improved |
| Add feature rectification (ablation) | ResNet-50 Energy detector · ImageNet-1k OOD average | FPR95: 58.16 -> 27.11 | Improved |
| Rectify Block1 only instead of using no rectification (ablation) | ResNet-50 Energy detector · ImageNet-1k OOD average | FPR95: 58.16 -> 59.08 | Hurt |
| Rectify Block2 only instead of using no rectification (ablation) | ResNet-50 Energy detector · ImageNet-1k OOD average | FPR95: 58.16 -> 57.62 | Improved |
| Rectify Block3 only instead of using no rectification (ablation) | ResNet-50 Energy detector · ImageNet-1k OOD average | FPR95: 58.16 -> 53.84 | Improved |
| Rectify all blocks rather than Block4 only (ablation) | ResNet-50 Energy detector · ImageNet-1k OOD average | FPR95: 27.11 -> 26.12 | Improved |
| Use a very small truncation threshold (ablation) | BATS · ImageNet-1k and CIFAR-100 OOD detection | FPR95 | Hurt |
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.
- ReAct — ReAct: Out-of-distribution Detection With Rectified Activations (NeurIPS 2021) · better on 11 of 11 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
- Exploring Channel-Aware Typical Features for Out-of-Distribution Detection (AAAI 2024 · similarity 0.92)
- Exploiting Discrepancy in Feature Statistic for Out-of-Distribution Detection (AAAI 2024 · similarity 0.91)
- Leveraging Perturbation Robustness to Enhance Out-of-Distribution Detection (CVPR 2025 · similarity 0.91)
- Out-of-Distribution Detection with Prototypical Outlier Proxy (AAAI 2025 · similarity 0.91)
- Expecting The Unexpected: Towards Broad Out-Of-Distribution Detection (NeurIPS 2024 · similarity 0.90)
- Image Background Serves as Good Proxy for Out-of-distribution Data (ICLR 2024 · similarity 0.90)
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.
