Anomaly Detection
Anomaly Detection is a binary classification identifying unusual or unexpected patterns in a dataset, which deviate significantly from the majority of the data. The goal of anomaly detection is to identify such anomalies, which could represent errors, fraud, or other types of unusual events, and flag them for further investigation.
[Image source]: GAN-based Anomaly Detection in Imbalance Problems
Papers
Showing 11–20 of 4856 papers
All datasetsMVTec ADVisAMVTec LOCO ADOne-class CIFAR-10CUHK AvenueShanghaiTechFishyscapes L&FUCR Anomaly ArchiveMPDDBTADOne-class CIFAR-100UBnormal
Benchmark Results
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | CPR-faster(TensorRT) | FPS | 1,016 | — | Unverified |
| 2 | CPR-fast(TensorRT) | FPS | 362 | — | Unverified |
| 3 | CPR(TensorRT) | FPS | 130 | — | Unverified |
| 4 | GLASS | Detection AUROC | 99.9 | — | Unverified |
| 5 | UniNet | Detection AUROC | 99.9 | — | Unverified |
| 6 | INP-Fomer ViT-L (model-unified multi-class) | Detection AUROC | 99.8 | — | Unverified |
| 7 | DDAD | Detection AUROC | 99.8 | — | Unverified |
| 8 | EfficientAD (early stopping) | Detection AUROC | 99.8 | — | Unverified |
| 9 | PBAS | Detection AUROC | 99.8 | — | Unverified |
| 10 | HETMM | Detection AUROC | 99.8 | — | Unverified |
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | UniNet | Detection AUROC | 99.8 | — | Unverified |
| 2 | GLAD | Detection AUROC | 99.5 | — | Unverified |
| 3 | UniNet(model-unified multi-class) | Detection AUROC | 99.15 | — | Unverified |
| 4 | DDAD | Detection AUROC | 98.9 | — | Unverified |
| 5 | Dinomaly ViT-L (model-unified multi-class) | Detection AUROC | 98.9 | — | Unverified |
| 6 | INP-Former ViT-B (model-unified multi-class) | Detection AUROC | 98.9 | — | Unverified |
| 7 | DiffusionAD | Detection AUROC | 98.8 | — | Unverified |
| 8 | GLASS | Detection AUROC | 98.8 | — | Unverified |
| 9 | TransFusion | Detection AUROC | 98.7 | — | Unverified |
| 10 | HETMM | Detection AUROC | 98.1 | — | Unverified |