| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Scene Graph Generation | Visual Genome (test) | R@5030.9 | 86 | |
| Scene Graph Classification | Visual Genome (test) | Recall@10049 | 63 | |
| Predicate Classification | Visual Genome | Recall@5081.1 | 54 | |
| Predicate Classification | Visual Genome (test) | R@5081.9 | 50 | |
| Scene Graph Classification | Visual Genome | R@5044.5 | 45 | |
| Predicate Classification | Visual Genome (VG) 150 object categories 50 relationship categories (test) | mR@10062.6 | 44 | |
| Layout-to-Image Synthesis | Visual Genome (VG) (test) | FID15.63 | 35 | |
| Scene Graph Detection | Visual Genome | Recall@10035.8 | 31 | |
| Scene Graph Detection | Visual Genome (VG) (test) | mR@5018.6 | 29 | |
| Multi-Label Classification | Visual Genome VG256 (test) | mAP50.9 | 24 | |
| Predicate Classification | Visual Genome 1.0 (test) | R@10075.2 | 22 | |
| Scene Graph Detection (SGDet) | Visual Genome (VG) | R@5033.5 | 21 | |
| Scene Graph Classification (SGCls) | Visual Genome (VG) | ng mR@5027.9 | 20 | |
| Layout-to-Image Generation | Visual Genome | FID0 | 20 | |
| Scene Graph Classification (SGCls) | Visual Genome | R@10042.3 | 19 | |
| Scene Graph Detection | Visual Genome (VG) Zero-Shot | R@50260 | 19 | |
| Scene Graph Classification | Visual Genome (VG) Zero-Shot | R@503.4 | 19 | |
| Predicate Classification | Visual Genome (VG) Zero-Shot | Recall@5014.4 | 19 | |
| Sentence-to-Graph Retrieval | Visual Genome Gallery Size 5000 (test) | R@205.2 | 19 | |
| Sentence-to-Graph Retrieval | Visual Genome Gallery Size 1000 (test) | Recall@2020.8 | 19 | |
| Region Captioning | Visual Genome | METEOR19.7 | 18 | |
| Scene Graph Generation | Visual Genome VG150 (test) | R@5032.9 | 16 | |
| Dense Captioning | Visual Genome | mAP16.2 | 16 | |
| Weakly Supervised Grounding | Visual Genome (VG) (test) | Accuracy (Pointing Game)55.91 | 15 | |
| Scene Graph Classification | Visual Genome (VG) | mR@10018.8 | 14 |