Video-to-Text Retrieval on DiDeMo (test)
67.5R@1VidLA
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| VidLABackbone=CLIP-ViT-B/16, dual-softmax=true2024.03 | 67.5 | 91.7 | — | — | 87.5 | 82.2 | |
| VidLABackbone=CLIP-ViT-B/32, dual-softmax=true2024.03 | 62.2 | 91.1 | — | — | 85.1 | 79.5 | |
| VidLABackbone=CLIP-ViT-B/162024.03 | 59.6 | 90.2 | — | — | 84.4 | 78 | |
| UMTBackbone=CLIP-ViT-B/16, two-stage method with candidate re-ranking=true2024.03 | 59.5 | 90.5 | — | — | 84.9 | 78.3 | |
| InternVideoBackbone=ViT-L, dual-softmax=true2024.03 | 59.1 | — | — | — | — | — | |
| InternVideo2 6Bretrieval_stage=Dual, model_scale=6B, zero_shot=true2026.01 | 57.1 | — | — | — | 79.9 | — | |
| VidLABackbone=CLIP-ViT-B/322024.03 | 55 | 88.9 | — | — | 82.8 | 75.6 | |
| VIRTUE v1 7Bretrieval_stage=Dual, model_scale=7B, zero_shot=true2026.01 | 52.5 | — | — | — | 79 | — | |
| DiffusionRetBackbone=CLIP-ViT-B/32, dual-softmax=true2024.03 | 50.3 | 82.9 | — | — | 75.1 | 69.4 | |
| DRLBackbone=CLIP-ViT-B/162024.03 | 49.9 | 83.3 | — | — | 75.4 | 69.5 | |
| NeighborRetr2025.03 | 48.4 | 83.7 | 2 | 9 | 74.6 | — | |
| PAU2023.09 | 48.1 | 85.7 | 2 | 9.8 | 74.2 | — | |
| X-CLIPBackbone=ViT-B/162022.07 | 47.8 | — | — | 10.5 | 76.8 | — | |
| RAPType=Adapter, Frozen visual encoder=true2024.05 | 47.7 | 83.2 | — | 9.5 | 74.4 | — | |
| CLIP4Clip-MeanPBackbone=ViT-B/162022.07 | 47.2 | — | — | 10.5 | 74 | — | |
| VIRTUE v2 7Bretrieval_stage=Dual, model_scale=7B, zero_shot=true2026.01 | 47 | — | — | — | 75.8 | — | |
| BaselineCondition=without PAU2023.09 | 46.9 | 84.2 | 2 | 8.4 | 74.4 | — | |
| HBI2023.03 | 46.2 | 82.7 | 2 | 8.7 | 73 | — | |
| Diffusion2025.03 | 46.2 | 82.2 | 2 | 10.7 | 74.3 | — | |
| HBI2025.03 | 46.2 | 82.7 | 2 | 8.7 | 73 | — | |
| UCOFIA2023.09 | 46 | 81.5 | 2 | 12.1 | 71.9 | — | |
| EMCL-Net2025.03 | 45.7 | 82.7 | 2 | 10.9 | 74.3 | — | |
| CAMOEBackbone=CLIP-ViT-B/32, dual-softmax=true2024.03 | 45.5 | — | — | — | 71.2 | — | |
| DRLBackbone=CLIP-ViT-B/322024.03 | 45.4 | 82.1 | — | — | 72.6 | 66.7 | |
| CLIP4Clip-seqTransfBackbone=ViT-B/162022.07 | 44.7 | — | — | 10.6 | 74 | — | |
| VoPF+PParams (M)=0.4 (0.328%)2022.11 | 44.7 | 81.1 | 2 | 9.9 | 71.2 | — | |
| VoPF+CParams (M)=14.1 (11.785%)2022.11 | 44.4 | 81.8 | 2 | 9.5 | 71.8 | — | |
| RAPType=Adapter, Frozen visual encoder=false2024.05 | 44 | 80.1 | — | 10.5 | 69.2 | — | |
| VoPFParams (M)=0.1 (0.103%)2022.11 | 43.5 | 81.4 | 2 | 9.8 | 70.9 | — | |
| CLIP4Clip-meanPAggregation Strategy=mean pooling2023.09 | 43.4 | 80.2 | 2 | 17.5 | 69.9 | — | |
| X-CLIPBackbone=ViT-B/322022.07 | 43.1 | — | — | 10.9 | 72.2 | — | |
| X-CLIP2023.09 | 43.1 | — | — | 10.9 | 72.2 | — | |
| X-Clip2025.03 | 43.1 | — | — | 10.9 | 77.2 | — | |
| CLIP4Clip-seqTransfAggregation Strategy=sequential Transformer2023.09 | 42.7 | 80.6 | 2 | 11.6 | 70.9 | — | |
| CLIP4Clip-meanPTraining Dataset=WIT + DiDeMo, End-to-End=true2021.04 | 42.5 | 80.2 | 2 | 11.6 | 70.6 | — | |
| CLIP4Clip-MeanPBackbone=ViT-B/322022.07 | 42.5 | — | — | 11.6 | 70.6 | — | |
| CLIP4Clip2023.09 | 42.5 | 80.2 | 2 | 11.6 | 70.6 | — | |
| CLIP4ClipBackbone=CLIP-ViT-B/322024.03 | 42.5 | 80.2 | — | — | 70.6 | 64.4 | |
| CLIP4Clip-seqLSTMTraining Dataset=WIT + DiDeMo, End-to-End=true2021.04 | 42.4 | 79.2 | 2 | 11.8 | 69.2 | — | |
| CLIP4ClipType=Fine-tune, Frozen visual encoder=false2024.05 | 42.2 | 79.3 | — | 11.8 | 70.3 | — | |
| VIRTUE-Embed 7Bretrieval_stage=Single, model_category=MLLM-based, model_scale=7B, zero_shot=true2026.01 | 42.2 | — | — | — | 68.5 | — | |
| CLIP4Clip-seqTransfTraining Dataset=WIT + DiDeMo, End-to-End=true2021.04 | 41.4 | 79.1 | 2 | 12.4 | 68.2 | — | |
| CLIP4Clip-seqTransfBackbone=ViT-B/322022.07 | 41.4 | — | — | 12.4 | 68.2 | — | |
| CLIP4Clip2023.03 | 41.4 | 79.1 | 2 | 12.4 | 68.2 | — | |
| VoPPParams (M)=0.5 (0.441%)2022.11 | 40.6 | 78.6 | 2 | 11.6 | 68.3 | — | |
| VoPPType=Prompt2024.05 | 40.6 | 78.6 | — | 11.6 | 68.3 | — | |
| FullParams (M)=119.8 (100%)2022.11 | 40.2 | 78.7 | 2 | 11.9 | 68.4 | — | |
| SSFType=Adapter2024.05 | 40 | 77.4 | — | 13.2 | 67.1 | — | |
| LORAType=Adapter2024.05 | 39.7 | 77.3 | — | 13.9 | 66.8 | — | |
| VoPCParams (M)=14.3 (11.898%)2022.11 | 39.1 | 76.7 | 3 | 13.8 | 65.3 | — | |
| VoPCType=Prompt2024.05 | 39.1 | 76.7 | — | 13.8 | 65.3 | — | |
| CaRe 7Bretrieval_stage=Single, model_category=MLLM-based, model_scale=7B, zero_shot=true2026.01 | 39.1 | — | — | — | 66 | — | |
| VoPParams (M)=0.1 (0.103%)2022.11 | 38.1 | 76.5 | 2 | 13.5 | 65.7 | — | |
| PartialParams (M)=7.7 (6.410%)2022.11 | 36.9 | 74.5 | 2 | 17 | 64.2 | — | |
| BiasParams (M)=0.1 (0.104%)2022.11 | 36.8 | 75.8 | 2 | 15.1 | 65.7 | — | |
| AdapterATTNParams (M)=2.0 (1.655%)2022.11 | 36.3 | 74.8 | 2 | 15.4 | 64.4 | — | |
| ST-AdapterType=Adapter2024.05 | 35.9 | 72 | — | 20.1 | 61 | — | |
| AdapterFFNParams (M)=2.0 (1.655%)2022.11 | 35.6 | 75.6 | 3 | 14.8 | 64.3 | — | |
| LanguageBindretrieval_stage=Single, model_category=CLIP-based, zero_shot=true2026.01 | 35.6 | — | — | — | 62.8 | — | |
| ProjParams (M)=0.7 (0.547%)2022.11 | 34.5 | 72.6 | 3 | 18.8 | 60.9 | — | |
| VPTType=Prompt2024.05 | 33.1 | 69.9 | — | 22.7 | 59.8 | — | |
| LaTzero-shot=false2022.07 | 32.7 | 72.7 | 3 | — | 61.1 | — | |
| FullTriplets (Ratio)=-, Time (mins)=-2026.04 | 32.7 | — | — | — | 58.6 | — | |
| CoOpType=Prompt2024.05 | 32.3 | 68.2 | — | 23.4 | 57 | — | |
| HoPATriplets (Ratio)=200 (2.38%), Time (mins)=45.72026.04 | 31.8 | — | — | — | 57.2 | — | |
| HoPATriplets (Ratio)=100 (1.19%), Time (mins)=41.92026.04 | 31.5 | — | — | — | 56.6 | — | |
| HoPATriplets (Ratio)=500 (5.96%), Time (mins)=53.62026.04 | 31.3 | — | — | — | 57.3 | — | |
| LoRSTriplets (Ratio)=500 (5.96%), Time (mins)=122.32026.04 | 30.3 | — | — | — | 56.6 | — | |
| RepBlendTriplets (Ratio)=500 (5.96%), Time (mins)=122.82026.04 | 30 | — | — | — | 56.1 | — | |
| RandomTriplets (Ratio)=500 (5.96%), Time (mins)=-2026.04 | 29.9 | — | — | — | 54.9 | — | |
| MTT-VLTriplets (Ratio)=500 (5.96%), Time (mins)=120.22026.04 | 29.5 | — | — | — | 56.3 | — | |
| LoRSTriplets (Ratio)=200 (2.38%), Time (mins)=96.32026.04 | 29.2 | — | — | — | 54.8 | — | |
| TESLATriplets (Ratio)=500 (5.96%), Time (mins)=118.92026.04 | 29.1 | — | — | — | 55.9 | — | |
| RepBlendTriplets (Ratio)=200 (2.38%), Time (mins)=97.42026.04 | 29 | — | — | — | 55 | — | |
| Long-CLIP L/14retrieval_stage=Single, model_category=CLIP-based, model_scale=L/14, zero_shot=true2026.01 | 28.5 | — | — | — | 54.1 | — | |
| RandomTriplets (Ratio)=200 (2.38%), Time (mins)=-2026.04 | 28.3 | — | — | — | 51.4 | — | |
| LoRSTriplets (Ratio)=100 (1.19%), Time (mins)=79.42026.04 | 28 | — | — | — | 52.7 | — | |
| MTT-VLTriplets (Ratio)=200 (2.38%), Time (mins)=95.02026.04 | 27.9 | — | — | — | 53.2 | — | |
| TESLATriplets (Ratio)=200 (2.38%), Time (mins)=97.42026.04 | 27.8 | — | — | — | 53.5 | — | |
| RepBlendTriplets (Ratio)=100 (1.19%), Time (mins)=78.22026.04 | 27.7 | — | — | — | 52.9 | — | |
| RandomTriplets (Ratio)=100 (1.19%), Time (mins)=-2026.04 | 27.4 | — | — | — | 50.3 | — | |
| MTT-VLTriplets (Ratio)=100 (1.19%), Time (mins)=77.32026.04 | 26.6 | — | — | — | 51.5 | — | |
| TESLATriplets (Ratio)=100 (1.19%), Time (mins)=78.12026.04 | 26.6 | — | — | — | 51.5 | — | |
| ATP2022.06 | 26.1 | — | — | — | 50.5 | — | |
| LaTzero-shot=true2022.07 | 22.5 | 56.8 | 7 | — | 45.2 | — | |
| CLIP4Clip-tightTransfTraining Dataset=WIT + DiDeMo, End-to-End=true2021.04 | 21.5 | 64.8 | 5 | 22.4 | 51.1 | — | |
| TT-CE+Training Dataset=DiDeMo, End-to-End=false2021.04 | 21.1 | 61.1 | 6.3 | — | 47.3 | — | |
| TT-CE+2022.07 | 21.1 | — | — | — | 47.3 | — | |
| TT-CE2023.03 | 21.1 | 61.1 | — | 6.3 | 47.3 | — | |
| TT-CE2023.09 | 21.1 | 61.1 | 6.3 | — | 47.3 | — | |
| ClipBERT2022.06 | 20.4 | — | — | — | 48 | — | |
| CLIPinput_format=single-frame2022.06 | 20.2 | — | — | — | 42.5 | — | |
| CLIP4ClipType=Fine-tune, Frozen visual encoder=true2024.05 | 20.2 | 55 | — | 43.1 | 44.2 | — | |
| MMLoss=MM2021.03 | 17 | 52.3 | 10 | 42.6 | — | — | |
| VideoCLIPzero-shot=true2022.06 | 16.6 | — | — | — | 46.9 | — | |
| MMLoss=MM [19]2021.03 | 15.6 | — | 8.2 | 42.4 | — | — | |
| CETraining Dataset=DiDeMo, End-to-End=false2021.04 | 15.6 | — | 8.2 | 42.4 | 40.9 | — | |
| CE2022.07 | 15.6 | — | — | 42.4 | 40.9 | — | |
| CE2023.03 | 15.6 | — | 8.2 | 42.4 | 40.9 | — | |
| CE2023.09 | 15.6 | — | 8.2 | 42.4 | 40.9 | — |