Grounded video description on ActivityNet Entities (test)
24.1B@1GVD
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| GVDModule=MHA + SCM-VG2019.05 | 24.1 | 2.63 | 11.4 | 49 | 15.1 | 7.81 | 27.1 | |
| GVDModule=MHA2019.05 | 23.6 | 2.35 | 11 | 45.5 | 14.9 | 7.59 | 25 | |
| GVDModule=MHA + SCM2019.05 | 23.6 | 2.54 | 11.2 | 47.7 | 15 | 7.3 | 24.4 | |
| GVDModule=VCM + SCM2019.05 | 23.1 | 2.34 | 10.9 | 46.1 | 14.5 | — | — | |
| Masked Transformer2019.05 | 22.9 | 2.41 | 10.6 | 46.1 | 13.7 | — | — | |
| Bi-LSTM+TempoAttn2019.05 | 22.8 | 2.17 | 10.2 | 42.2 | 11.8 | — | — |