Dense Video Object Captioning on VidSTG (test)
43.9CapAOW-VISCapTor+CAROQ [18] (online)
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| OW-VISCapTor+CAROQ [18] (online)Mode=online2024.04 | 43.9 | 53.1 | 60.1 | 54 | 62.6 | |
| OW-VISCapTor+DEVA [7] (online)Mode=online2024.04 | 40.1 | 51.7 | 60 | 56.3 | 63 | |
| DVOC-DS (joint training) [6]Mode=offline2024.04 | 36.8 | 51.6 | 65.5 | 56.9 | 69.3 | |
| Mask2Former [13]+DEVA [7]+ BLIP2 [11]Mode=online2024.04 | 34 | 48.5 | 59.6 | 56.4 | 60.1 | |
| DVOC-DS (disjoint training) [6]Mode=offline2024.04 | 10 | 28 | 45.9 | 48 | 39.8 |