Metric: AP (higher is better)
| # | Model↕ | AP▼ | Augmentations | Paper | Date↕ | Code |
|---|---|---|---|---|---|---|
| 1 | MaxViT-B | 53.4 | No | MaxViT: Multi-Axis Vision Transformer | 2022-04-04 | Code |
| 2 | MaxViT-S | 53.1 | No | MaxViT: Multi-Axis Vision Transformer | 2022-04-04 | Code |
| 3 | MaxViT-T | 52.1 | No | MaxViT: Multi-Axis Vision Transformer | 2022-04-04 | Code |
| 4 | DAT-S++ | 50.2 | No | DAT++: Spatially Dynamic Vision Transformer with... | 2023-09-04 | Code |
| 5 | DAT-T++ | 49.2 | No | DAT++: Spatially Dynamic Vision Transformer with... | 2023-09-04 | Code |
| 6 | DyHead (SAP) | 42.1 | No | Stochastic Subsampling With Average Pooling | 2024-09-25 | - |
| 7 | Faster R-CNN (ideal number of groups) | 40.7 | No | On the Ideal Number of Groups for Isometric Grad... | 2023-02-07 | - |
| 8 | DETReg (ours) | 30 | No | DETReg: Unsupervised Pretraining with Region Pri... | 2021-06-08 | Code |