Video Paragraph Captioning on ActivityNet Captions
11.74BLEU@4TDPC* (w/o RL)
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| TDPC* (w/o RL)Anno. (boundary annotations during inference)=No, Extra input modalities (e.g., flow, object detection)=Yes2023.03 | 11.74 | 15.64 | 26.55 | |
| GVLAnno. (boundary annotations during inference)=No, Extra input modalities (e.g., flow, object detection)=No2023.03 | 11.7 | 16.35 | 26.03 | |
| GVD*Anno. (boundary annotations during inference)=Yes, Extra input modalities (e.g., flow, object detection)=Yes2023.03 | 11.04 | 15.71 | 21.95 | |
| COOT*Anno. (boundary annotations during inference)=Yes, Extra input modalities (e.g., flow, object detection)=Yes2023.03 | 10.85 | 15.99 | 28.19 | |
| PDVCAnno. (boundary annotations during inference)=No, Extra input modalities (e.g., flow, object detection)=No2023.03 | 10.46 | 16.42 | 20.91 | |
| MART*Anno. (boundary annotations during inference)=Yes, Extra input modalities (e.g., flow, object detection)=Yes2023.03 | 10.33 | 15.68 | 23.42 | |
| MFT*Anno. (boundary annotations during inference)=No, Extra input modalities (e.g., flow, object detection)=Yes2023.03 | 10.29 | 14.73 | 19.12 | |
| AdvInf*Anno. (boundary annotations during inference)=Yes, Extra input modalities (e.g., flow, object detection)=Yes2023.03 | 10.04 | 16.6 | 20.97 | |
| HSEAnno. (boundary annotations during inference)=Yes, Extra input modalities (e.g., flow, object detection)=No2023.03 | 9.84 | 13.78 | 18.78 |