Research map / Label-efficient and robust learning / OOD and anomaly detection
Explainable Deep One-Class Classification
Read the paperCheck 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 FCDD's fully convolutional one-class classifier and spatial anomaly heatmap | FCDD · CIFAR-10 one-vs-rest | AUC: 0.95 vs 0.96 | Mixed |
| Use FCDD's fully convolutional one-class classifier and spatial anomaly heatmap | FCDD · ImageNet one-vs-rest | AUC: 0.94 vs 0.97 | Hurt |
| Use FCDD for unsupervised defect localization | MVTec-AD · MVTec-AD unsupervised | pixel-wise mean AUC: 0.92 vs 0.89 | Helped |
| Add a few labeled anomalies and train with their ground-truth maps | FCDD · MVTec-AD | pixel-wise mean AUC: 0.92 -> 0.96 | Helped |
| Use a smaller receptive field (53 rather than 243) (ablation) | FCDD · MVTec-AD | pixel-wise mean AUC: 0.75 -> 0.88 | Helped |
| Increase receptive field from 18 to 32 (ablation) | FCDD · CIFAR-10 | AUC: 0.9328 -> 0.9235 | Hurt |
| Increase Gaussian upsampling sigma from 4 to 14 (ablation) | FCDD · MVTec-AD | pixel-wise mean AUC: 0.8567 -> 0.9217 | Helped |
| Use grayscaled CIFAR-100 instead of EMNIST as outlier exposure data | FCDD · Fashion-MNIST | AUC: ~3 AUC percent points | Helped |
Closest papers by meaning
- AnomalyNCD: Towards Novel Anomaly Class Discovery in Industrial Scenarios (CVPR 2025 · similarity 0.84)
- Dinomaly: The Less Is More Philosophy in Multi-Class Unsupervised Anomaly Detection (CVPR 2025 · similarity 0.83)
- A Diffusion-Based Framework for Multi-Class Anomaly Detection (AAAI 2024 · similarity 0.83)
- ResAD: A Simple Framework for Class Generalizable Anomaly Detection (NeurIPS 2024 · similarity 0.83)
- CNC: Cross-modal Normality Constraint for Unsupervised Multi-class Anomaly Detection (AAAI 2025 · similarity 0.83)
- Detecting and Corrupting Convolution-based Unlearnable Examples (AAAI 2025 · similarity 0.83)
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.