Metric: mAP@0.5 (higher is better)
| # | Model↕ | mAP@0.5▼ | Augmentations | Paper | Date↕ | Code |
|---|---|---|---|---|---|---|
| 1 | PVT (Pyramid Vision Transformer; trained on PeopleArt and PopArt) | 80.5 | Yes | Poses of People in Art: A Data Set for Human Pos... | 2023-01-12 | - |
| 2 | TOOD (Task-aligned One-stage Object Detection; trained on PeopleArt and PoPArt) | 78 | Yes | Poses of People in Art: A Data Set for Human Pos... | 2023-01-12 | - |
| 3 | PVT (Pyramid Vision Transformer) | 76 | No | Poses of People in Art: A Data Set for Human Pos... | 2023-01-12 | - |
| 4 | TOOD (Task-aligned One-stage Object Detection) | 75 | No | Poses of People in Art: A Data Set for Human Pos... | 2023-01-12 | - |
| 5 | FasterRCNN (trained on StyleCOCO) | 68 | No | Improving Object Detection in Art Images Using O... | 2021-02-12 | Code |
| 6 | Fast R-CNN | 59 | No | Detecting People in Artwork with CNNs | 2016-10-27 | Code |