Object Detection on VOC 2007
86.46mAPCLOP
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| CLOPPre-trained=ImageNet, Backbone=ResNet-50 (4x), Evaluation Protocol=fine-tune2025.11 | 86.46 | — | — | |
| FPModel=YOLO V5-S, Optimization Technique=Full Precision, Evaluation Protocol=Fine-tuning2025.03 | 85.8 | — | — | |
| CCSSLPre-trained=ImageNet, Backbone=ResNet-50 (4x), Evaluation Protocol=fine-tune2025.11 | 85.8 | — | — | |
| SimMatchV2Pre-trained=ImageNet, Backbone=ResNet-50 (4x), Evaluation Protocol=fine-tune2025.11 | 85.46 | — | — | |
| SupConPre-trained=ImageNet, Backbone=ResNet-50 (4x), Evaluation Protocol=fine-tune2025.11 | 85.17 | — | — | |
| HOTModel=YOLO V5-S, Optimization Technique=HOT, Evaluation Protocol=Fine-tuning2025.03 | 85.1 | — | — | |
| SsCLPre-trained=ImageNet, Backbone=ResNet-50 (4x), Evaluation Protocol=fine-tune2025.11 | 85.04 | — | — | |
| LBP-WHTModel=YOLO V5-S, Optimization Technique=LBP-WHT, Evaluation Protocol=Fine-tuning2025.03 | 85 | — | — | |
| SimCLRPre-trained=ImageNet, Backbone=ResNet-50 (4x), Evaluation Protocol=fine-tune2025.11 | 84.1 | — | — | |
| SimMatchPre-trained=ImageNet, Backbone=ResNet-50 (4x), Evaluation Protocol=fine-tune2025.11 | 83.93 | — | — | |
| FixMatchPre-trained=ImageNet, Backbone=ResNet-50 (4x), Evaluation Protocol=fine-tune2025.11 | 83.72 | — | — | |
| ISyNet-N32021.09 | 80.9 | — | — | |
| ResNet-34+2021.09 | 79.67 | — | — | |
| ISyNet-N1-S32021.09 | 79.13 | — | — | |
| ISyNet-N1-S22021.09 | 78.94 | — | — | |
| ISyNet-N1-S12021.09 | 78.4 | — | — | |
| ISyNet-N22021.09 | 78.3 | — | — | |
| ResNet-50+2021.09 | 78.18 | — | — | |
| ISyNet-N12021.09 | 78 | — | — | |
| ResNet-18+2021.09 | 76.54 | — | — | |
| ISyNet-N02021.09 | 76 | — | — | |
| DETRModel=DETR2026.05 | 72.2 | — | — | |
| WWT-DETModel=WWT, Task head setting=Proposed task-head2026.05 | 63.1 | — | — | |
| WWT + det. head on tokensModel=WWT, Task head setting=heads-on-patch-tokens2026.05 | 37.8 | — | — | |
| RetinaNet+DALBackbone=RetinaNet+DAL2020.12 | 0.761 | 0.761 | — | |
| RetinaNetBackbone=RetinaNet2020.12 | 0.749 | 0.749 | — | |
| Gao et al.Iterations x Batch size=150K×64, Open-vocabulary=true2022.07 | — | — | 59.2 | |
| VL-PLMIterations x Batch size=180K×16, Open-vocabulary=true2022.07 | — | — | 67.4 |