Scene Graph Anticipation on Action Genome (test)
38.8R@10HyperGLM
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| HyperGLMVideo input fraction (F)=0.9, Pre-trained detector=Faster R-CNN2024.11 | 38.8 | 22.3 | 51.5 | 33 | 65.2 | 48.6 | |
| HyperGraphVideo input fraction (F)=0.9, Pre-trained detector=Faster R-CNN2024.11 | 37.5 | 19.1 | 49.3 | 31.4 | 62.3 | 47.4 | |
| SceneSayerSDEVideo input fraction (F)=0.9, Pre-trained detector=Faster R-CNN2024.11 | 37.3 | 20.8 | 48.6 | 30.9 | 61.6 | 46.8 | |
| SceneSayerODEVideo input fraction (F)=0.9, Pre-trained detector=Faster R-CNN2024.11 | 36.6 | 17.8 | 48.3 | 27.4 | 61.3 | 43.4 | |
| DSGDetr++Video input fraction (F)=0.9, Pre-trained detector=Faster R-CNN2024.11 | 22.2 | 11.4 | 37.1 | 21 | 61 | 39.5 | |
| STTran++Video input fraction (F)=0.9, Pre-trained detector=Faster R-CNN2024.11 | 20.2 | 8.9 | 35.7 | 18.4 | 60.2 | 38.8 | |
| DSGDetr+Video input fraction (F)=0.9, Pre-trained detector=Faster R-CNN2024.11 | 17.9 | 4.7 | 27.7 | 9.7 | 51.4 | 25.9 | |
| STTran+Video input fraction (F)=0.9, Pre-trained detector=Faster R-CNN2024.11 | 17.5 | 4.6 | 26.8 | 9.2 | 49.6 | 24.3 |