Anomaly Detection and Localization on MVTecLOCO (test)
84.1I-AUROCULSAD
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| ULSADThroughput (img/s)=33.42, GPU Memory (GB)=2.17, Batch size=32, Hardware (GPU)=NVIDIA A100, Representation=2 layers of pre-trained network2024.10 | 84.1 | 80.06 | 73.73 | |
| PatchCoreThroughput (img/s)=32.7, GPU Memory (GB)=6.8, Batch size=32, Hardware (GPU)=NVIDIA A1002024.10 | 81.49 | 75.77 | 69.09 | |
| EffADThroughput (img/s)=23.33, GPU Memory (GB)=3.48, Batch size=32, Hardware (GPU)=NVIDIA A1002024.10 | 80.62 | 70.36 | 66.96 | |
| CFLOWThroughput (img/s)=11.69, GPU Memory (GB)=2.57, Batch size=32, Hardware (GPU)=NVIDIA A1002024.10 | 73.62 | 76.99 | 66.93 | |
| DFRThroughput (img/s)=15.08, GPU Memory (GB)=5.85, Batch size=32, Hardware (GPU)=NVIDIA A100, Representation=multiscale (12 layers of pre-trained network)2024.10 | 72.87 | 61.72 | 69.78 | |
| FastFlowThroughput (img/s)=30.21, GPU Memory (GB)=1.69, Batch size=32, Hardware (GPU)=NVIDIA A1002024.10 | 71 | 75.55 | 53.04 | |
| PaDiMThroughput (img/s)=33.45, GPU Memory (GB)=1.92, Batch size=32, Hardware (GPU)=NVIDIA A1002024.10 | 68.38 | 71.32 | 67.97 | |
| DRÆMThroughput (img/s)=10.06, GPU Memory (GB)=7.95, Batch size=32, Hardware (GPU)=NVIDIA A1002024.10 | 62.35 | 63.69 | 40.06 | |
| RDThroughput (img/s)=34.87, GPU Memory (GB)=1.93, Batch size=32, Hardware (GPU)=NVIDIA A1002024.10 | 61.56 | 68.55 | 66.28 |