Anomaly Detection On Visa
평가 지표
Detection AUROC
평가 결과
이 벤치마크에서 각 모델의 성능 결과
비교 표
| 모델 이름 | Detection AUROC |
|---|---|
| cflow-ad-real-time-unsupervised-anomaly | 91.5 |
| anomalydino-boosting-patch-based-few-shot | 97.6 |
| winclip-zero-few-shot-anomaly-classification | 84.6 |
| anomaly-detection-via-reverse-distillation | - |
| a-unified-anomaly-synthesis-strategy-with | 98.8 |
| musc-zero-shot-industrial-anomaly | 92.8 |
| anomalyclip-object-agnostic-prompt-learning | 82.1 |
| anomalydino-boosting-patch-based-few-shot | 92.6 |
| a-zero-few-shot-anomaly-classification-and | 78.0 |
| asymmetric-student-teacher-networks-for | 94.9 |
| spot-the-difference-self-supervised-pre | 87.8 |
| student-teacher-feature-pyramid-matching-for | 83.3 |
| segment-any-anomaly-without-training-via | - |
| 모델 14 | 97.5 |
| winclip-zero-few-shot-anomaly-classification | 78.1 |
| efficientad-accurate-visual-anomaly-detection | 97.5 |
| padim-a-patch-distribution-modeling-framework | - |
| anoddpm-anomaly-detection-with-denoising | 78.2 |
| sub-image-anomaly-detection-with-deep-pyramid | 82.1 |
| diffusionad-denoising-diffusion-for-anomaly | 98.8 |
| d3ad-dynamic-denoising-diffusion | 96.0 |
| draem-a-discriminatively-trained | - |
| efficientad-accurate-visual-anomaly-detection | 98.1 |
| anomalydino-boosting-patch-based-few-shot | 89.7 |
| adaclip-adapting-clip-with-hybrid-learnable | 85.8 |
| progressive-boundary-guided-anomaly-synthesis | 97.7 |
| vcp-clip-a-visual-context-prompting-model-for | - |
| fair-frequency-aware-image-restoration-for | 97.1 |
| unlocking-the-potential-of-reverse | 96.5 |
| uninformed-students-student-teacher-anomaly | 93.2 |
| transfusion-a-transparency-based-diffusion | 98.7 |
| glad-towards-better-reconstruction-with | 99.5 |
| anomaly-detection-with-conditioned-denoising | 98.9 |
| realnet-a-feature-selection-network-with | 97.8 |
| exploring-intrinsic-normal-prototypes-within | 98.9 |
| dsr-a-dual-subspace-re-projection-network-for | - |
| beyond-dents-and-scratches-logical | 89.1 |
| 2408-03143 | 93.4 |
| reconstruction-from-edge-image-combined-with | 94.2 |
| winclip-zero-few-shot-anomaly-classification | 83.8 |
| fastflow-unsupervised-anomaly-detection-and | - |
| cfa-coupled-hypersphere-based-feature | 92.0 |
| anomalydino-boosting-patch-based-few-shot | 87.4 |
| anomaly-localization-by-modeling-perceptual | - |
| dinomaly-the-less-is-more-philosophy-in-multi | 98.9 |
| hard-nominal-example-aware-template-mutual | 98.1 |
| winclip-zero-few-shot-anomaly-classification | 87.3 |