Text-to-Video Retrieval on MSR-VTT 1k-A (test)
48.5R@1MDMMT-2
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| MDMMT-2Configuration=1k-A2022.03 | 48.5 | 75.4 | 83.9 | 2 | 13.8 | |
| CLIP2TV2022.03 | 48.3 | 74.6 | 82.8 | 2 | 14.9 | |
| Query-scoringagg.=Q-score, #agg. params=12022.05 | 47.7 | 74.1 | 82.9 | 2 | 11.5 | |
| CLIP2VideoRefinement=QB-Norm2022.03 | 47.2 | 73 | 83 | 2 | — | |
| MDMMT-2Configuration=full2022.03 | 46.5 | 74.3 | 83.3 | 2 | 14.1 | |
| LAFF2022.03 | 45.8 | 71.5 | 82 | — | — | |
| Clip2Videoagg.=TDB, TAB, #agg. params=19M2022.05 | 45.6 | 72.6 | 81.7 | 2 | 14.6 | |
| CLIP2Video2022.03 | 45.6 | 72.6 | 81.7 | 2 | 14.6 | |
| CAMOEagg.=S.E attn.2022.05 | 44.6 | 72.6 | 81.8 | 2 | 13.3 | |
| CAMoE2022.03 | 44.6 | 72.6 | 81.8 | 2 | 13.3 | |
| Clip4clipagg.=seqTransf, #agg. params=4M2022.05 | 44.5 | 71.4 | 81.6 | 2 | 15.3 | |
| CLIP4Clip2022.03 | 44.5 | 71.4 | 81.6 | 2 | 15.3 | |
| Clip4clipagg.=mean, #agg. params=02022.05 | 43.1 | 70.4 | 80.8 | 2 | 16.2 | |
| Clip4clipagg.=tight Transf, #agg. params=4M2022.05 | 40.2 | 71.5 | 70.5 | 2 | 13.4 | |
| MDMMT2022.03 | 38.9 | 69 | 79.7 | 2 | 16.5 | |
| HD-VILAprotocol=fine-tuning2021.11 | 35.6 | 65.3 | 78 | 3 | — | |
| LaTZero-shot=false2022.07 | 35.3 | 61.3 | 72.9 | 3 | — | |
| Frozen2022.05 | 32.5 | 61.5 | 71.2 | 3 | — | |
| CLIP aggAggregation=agg2022.03 | 31.2 | 53.7 | 64.2 | 4 | — | |
| FrozenZero-shot=false2022.07 | 31 | 59.5 | 70.5 | 3 | — | |
| VideoCLIPprotocol=fine-tuning2021.11 | 30.9 | 55.4 | 66.8 | — | — | |
| HiTZero-shot=false2022.07 | 30.7 | 60.9 | 73.2 | 3 | — | |
| SSB2022.05 | 30.1 | 58.5 | 69.3 | 3 | — | |
| Support Setprotocol=fine-tuning2021.11 | 30.1 | 58.5 | 69.3 | 3 | — | |
| Support-SetZero-shot=false2022.07 | 30.1 | 58.5 | 69.3 | 3 | — | |
| SSB2022.03 | 30.1 | 58.5 | 69.3 | 3 | — | |
| TeachText2022.05 | 29.6 | 61.6 | 74.2 | 3 | — | |
| VLMprotocol=fine-tuning2021.11 | 28.1 | 55.5 | 67.4 | 4 | — | |
| AVLnet2022.03 | 27.1 | 55.6 | 66.6 | 4 | — | |
| MMT2022.05 | 26.6 | 57.1 | 69.6 | 4 | — | |
| MMTprotocol=fine-tuning2021.11 | 26.6 | 57.1 | 69.6 | 4 | — | |
| MMTZero-shot=false2022.07 | 26.6 | 57.1 | 69.6 | 4 | — | |
| MMT2022.03 | 26.6 | 57.1 | 69.6 | 4 | 24 | |
| LaTZero-shot=true2022.07 | 23.4 | 44.1 | 53.3 | 8 | — | |
| CLIP2022.03 | 22.5 | 44.3 | 53.7 | 8 | 61.7 | |
| ClipBERTprotocol=fine-tuning2021.11 | 22 | 46.8 | 59.9 | 6 | — | |
| ClipBERTZero-shot=false2022.07 | 22 | 46.8 | 59.9 | 6 | — | |
| CE2022.05 | 20.9 | 48.8 | 62.4 | 6 | — | |
| CEprotocol=fine-tuning2021.11 | 20.9 | 48.8 | 62.4 | 6 | — | |
| CEZero-shot=false2022.07 | 20.9 | 48.8 | 62.4 | 6 | — | |
| CE2022.03 | 20.9 | 48.8 | 62.4 | 6 | 28.2 | |
| FrozenZero-shot=true2022.07 | 18.7 | 39.5 | 51.6 | 10 | — | |
| DECEMBERTprotocol=fine-tuning2021.11 | 17.5 | 44.3 | 58.6 | 9 | — | |
| Noise-EstimationZero-shot=false2022.07 | 17.4 | 41.6 | 53.6 | 8 | — | |
| HEROprotocol=fine-tuning2021.11 | 16.8 | 43.4 | 57.7 | — | — | |
| ActBertZero-shot=false2022.07 | 16.3 | 42.8 | 56.9 | 10 | — | |
| HowTo100Mprotocol=fine-tuning2021.11 | 14.9 | 40.2 | 52.8 | 9 | — | |
| HT2022.03 | 14.9 | 40.2 | 52.8 | 9 | — | |
| HD-VILAprotocol=zero-shot2021.11 | 14.6 | 34.4 | 44.1 | 15 | — | |
| Support-SetZero-shot=true2022.07 | 12.7 | 27.5 | 36.2 | 24 | — | |
| HowTo100MZero-shot=false2022.07 | 12.6 | 36.2 | 48.1 | 13 | — | |
| VideoCLIPprotocol=zero-shot2021.11 | 10.4 | 22.2 | 30 | — | — | |
| JSFusion2022.05 | 10.2 | 31.2 | 43.2 | 12 | — | |
| JSFusion2022.03 | 10.2 | 31.2 | 43.2 | 13 | — | |
| HT MIL-NCEprotocol=zero-shot2021.11 | 9.9 | 24 | 32.4 | 29.5 | — | |
| E2E2022.03 | 9.9 | 24 | 32.4 | 29.5 | — | |
| Support Setprotocol=zero-shot2021.11 | 8.7 | 23 | 31.1 | 31 | — |