Text-Video Retrieval on MSR-VTT (split2)
24.8R@1TACo
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| TACoEvaluation Protocol=Finetuned, Video Backbone=S3D-HM2021.08 | 24.8 | 52.1 | 64 | 5 | |
| UniVL(v3)Evaluation Protocol=Finetuned, Video Backbone=S3D-HM2021.08 | 21.2 | 49.6 | 63.1 | 6 | |
| TACoLang.=BERT, Video=R-152+I-1012021.08 | 19.2 | 44.7 | 57.2 | 7 | |
| TJVEEvaluation Protocol=Finetuned, Video Backbone=R-152+I3D-X1012021.08 | 14.9 | 40.2 | 52.8 | 9 | |
| UniVL(v1)*Lang.=BERT, Video=R-152+I-1012021.08 | 14.6 | 39 | 52.6 | 10 | |
| JPoSELang.=w2v, Video=TSN+Flow2021.08 | 14.3 | 38.1 | 53 | 9 | |
| TVJELang.=w2v, Video=R-152+I-1012021.08 | 12.1 | 35 | 48 | 12 | |
| JSFusionLang.=BiLSTM, Video=R-1522021.08 | 10.2 | 31.2 | 43.2 | 13 | |
| MIL-NCEEvaluation Protocol=Zero-shot, Video Backbone=S3D-HM2021.08 | 9.9 | 24 | 32.4 | 29.5 | |
| TACoEvaluation Protocol=Zero-shot, Video Backbone=S3D-HM2021.08 | 9.8 | 25 | 33.4 | 29 | |
| ActBERTEvaluation Protocol=Zero-shot, Video Backbone=O-101+ R(2+1)D2021.08 | 8.6 | 23.4 | 33.1 | 36 | |
| TJVEEvaluation Protocol=Zero-shot, Video Backbone=R-152+I-1012021.08 | 7.5 | 21.2 | 29.6 | 38 | |
| RandomLang.=-, Video=-2021.08 | 0.1 | 0.5 | 1 | 500 |