HOI anticipation on VidHOI 2021 (val)
11.3mAP Fullspatio-temporal transformer network
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| spatio-temporal transformer networktau_a=1, Object Detector=YOLOv52023.06 | 11.3 | 52.53 | 43.61 | 28.81 | 35.86 | |
| spatio-temporal transformer networktau_a=3, Object Detector=YOLOv52023.06 | 10.65 | 51.63 | 43.6 | 28.66 | 35.37 | |
| spatio-temporal transformer networktau_a=5, Object Detector=YOLOv52023.06 | 10.19 | 51.69 | 42.49 | 28.22 | 34.88 | |
| spatio-temporal transformer networktau_a=7, Object Detector=YOLOv52023.06 | 10.14 | 50.72 | 42.1 | 27.6 | 34.14 | |
| STTrantau_a=1, Object Detector=YOLOv52023.06 | 8.8 | 53.31 | 27.62 | 18.85 | 27.15 | |
| STTrantau_a=7, Object Detector=YOLOv52023.06 | 8.75 | 52.18 | 26.35 | 18.01 | 25.9 | |
| STTrantau_a=5, Object Detector=YOLOv52023.06 | 8.67 | 52.96 | 26.97 | 18.48 | 26.54 | |
| STTrantau_a=3, Object Detector=YOLOv52023.06 | 8.32 | 52.58 | 26.99 | 18.41 | 26.48 |