Research Article
Multimodal Feature Learning for Video Captioning
Table 2
Comparison of different feature sets on MSVD dataset.
| | Feature sets | B@1 | B@2 | B@3 | B@4 | CIDEr |
| | CGN | 66.1 | 47.8 | 37.1 | 26.5 | 26.4 | | DSN + CGN | 76.0 | 58.1 | 45.7 | 35.8 | 50.0 | | SSN + CGN | 78.8 | 63.4 | 51.4 | 41.4 | 77.8 | | DSN + SSN + CGN | 84.8 | 70.8 | 60.0 | 50.0 | 94.3 |
|
|
