Research map / Label-efficient and robust learning / OOD and anomaly detection
SimpleNet: A Simple Network for Image Anomaly Detection and Localization
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 a target-domain feature adaptor and Gaussian-noise feature synthesis with a simple discriminator | SimpleNet · MVTec AD | I-AUROC: 99.1 -> 99.6 | Helped |
| Use a target-domain feature adaptor and Gaussian-noise feature synthesis with a simple discriminator | SimpleNet · CIFAR-10 one-class novelty detection | I-AUROC: 86.5 vs 86.5 | No clear effect |
| Use levels 2+3 instead of level 3 alone (ablation) | SimpleNet · MVTec AD | I-AUROC: 99.2 -> 99.6 | Helped |
| Add level 1 to levels 2+3 (ablation) | SimpleNet · MVTec AD | I-AUROC: 99.6 -> 99.1 | Hurt |
| Replace the linear feature adaptor with a nonlinear 1-layer MLP (ablation) | SimpleNet · MVTec AD | I-AUROC: 99.6 -> 98.3 | Hurt |
| Remove the feature adaptor (ablation) | SimpleNet · MVTec AD | I-AUROC: 99.6 -> 99.2 | Hurt |
| Replace truncated L1 loss with cross-entropy loss (ablation) | SimpleNet · MVTec AD | I-AUROC: 99.6 -> 99.4 | Hurt |
| Replace truncated L1 loss with cross-entropy loss (ablation) | SimpleNet · MVTec AD | P-AUROC: 98.1 -> 97.8 | Hurt |
| Use ResNet18 instead of WideResNet50 (ablation) | SimpleNet · MVTec AD | P-AUROC: 98.1 -> 95.7 | Hurt |
| Use ResNet50 instead of WideResNet50 (ablation) | SimpleNet · MVTec AD | I-AUROC: 99.6 vs 99.6 | No clear effect |
| Use ResNet50 instead of WideResNet50 (ablation) | SimpleNet · MVTec AD | P-AUROC: 98.1 -> 98.0 | Hurt |
Closest papers by meaning
- RealNet: A Feature Selection Network with Realistic Synthetic Anomaly for Anomaly Detection (CVPR 2024 · similarity 0.89)
- One-to-Normal: Anomaly Personalization for Few-shot Anomaly Detection (NeurIPS 2024 · similarity 0.88)
- AnomalyNCD: Towards Novel Anomaly Class Discovery in Industrial Scenarios (CVPR 2025 · similarity 0.87)
- Correcting Deviations from Normality: A Reformulated Diffusion Model for Multi-Class Unsupervised Anomaly Detection (CVPR 2025 · similarity 0.87)
- ResAD: A Simple Framework for Class Generalizable Anomaly Detection (NeurIPS 2024 · similarity 0.87)
- CostFilter-AD: Enhancing Anomaly Detection through Matching Cost Filtering (ICML 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.