Research map / Label-efficient and robust learning / Knowledge distillation
Anomaly Detection via Reverse Distillation From One-Class Embedding
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use reverse distillation with OCBE | MVTec anomaly detection · MVTec | AD AUROC: 96.1 -> 98.5 | Helped |
| Add OCE to the pre-trained bottleneck (ablation) | Reverse Distillation · MVTec | AD AUROC: 96.0 -> 97.9 | Helped |
| Add OCE to the pre-trained bottleneck (ablation) | Reverse Distillation · MVTec | AL AUROC: 96.9 -> 97.4 | Helped |
| Add OCE to the pre-trained bottleneck (ablation) | Reverse Distillation · MVTec | AL PRO: 91.2 -> 92.4 | Helped |
| Add MFF to Pre+OCE (ablation) | Reverse Distillation · MVTec | AD AUROC: 97.9 -> 98.5 | Helped |
| Add MFF to Pre+OCE (ablation) | Reverse Distillation · MVTec | AL AUROC: 97.4 -> 97.8 | Helped |
| Add MFF to Pre+OCE (ablation) | Reverse Distillation · MVTec | AL PRO: 92.4 -> 93.9 | Helped |
| Replace WideResNet50 teacher with ResNet18 (ablation) | Reverse Distillation · MVTec | AD AUROC: 98.5 -> 97.9 | Hurt |
| Replace WideResNet50 teacher with ResNet50 (ablation) | Reverse Distillation · MVTec | AD AUROC: 98.5 -> 98.4 | Hurt |
| Use only second-layer features instead of multi-scale M¹,²,³ (ablation) | Reverse Distillation · MVTec | AD AUROC: 98.5 -> 97.5 | Hurt |
| Use second- and third-layer features instead of multi-scale M¹,²,³ (ablation) | Reverse Distillation · MVTec | AD AUROC: 98.5 -> 98.0 | Hurt |
| Use second- and third-layer features instead of only third-layer features (ablation) | Reverse Distillation · MVTec | AD AUROC: 97.2 -> 98.0 | Helped |
Closest papers by meaning
- Unlocking the Potential of Reverse Distillation for Anomaly Detection (AAAI 2025 · similarity 0.95)
- Co-Progression Knowledge Distillation with Knowledge Prototype for Industrial Anomaly Detection (AAAI 2025 · similarity 0.89)
- ResAD: A Simple Framework for Class Generalizable Anomaly Detection (NeurIPS 2024 · similarity 0.88)
- AnomalyNCD: Towards Novel Anomaly Class Discovery in Industrial Scenarios (CVPR 2025 · similarity 0.88)
- ADPretrain: Advancing Industrial Anomaly Detection via Anomaly Representation Pretraining (NeurIPS 2025 · similarity 0.88)
- CNC: Cross-modal Normality Constraint for Unsupervised Multi-class Anomaly Detection (AAAI 2025 · similarity 0.87)
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.