Metric: video-to-text R@10 (higher is better)
| # | Model↕ | video-to-text R@10▼ | Extra Data | Paper | Date↕ | Code |
|---|---|---|---|---|---|---|
| 1 | HunYuan_tvr (huge) | 91.8 | Yes | Tencent Text-Video Retrieval: Hierarchical Cross... | 2022-04-07 | - |
| 2 | UMT-L (ViT-L/16) | 71.5 | Yes | Unmasked Teacher: Towards Training-Efficient Vid... | 2023-03-28 | Code |
| 3 | vid-TLDR (UMT-L) | 63.6 | Yes | vid-TLDR: Training Free Token merging for Light-... | 2024-03-20 | Code |
| 4 | CenterCLIP (ViT-B/16) | 55.8 | Yes | CenterCLIP: Token Clustering for Efficient Text-... | 2022-05-02 | Code |
| 5 | HunYuan_tvr | 55.7 | Yes | Tencent Text-Video Retrieval: Hierarchical Cross... | 2022-04-07 | - |
| 6 | EMCL-Net++ | 54.4 | No | Expectation-Maximization Contrastive Learning fo... | 2022-11-21 | Code |
| 7 | DiffusionRet | 51.5 | No | DiffusionRet: Generative Text-Video Retrieval wi... | 2023-03-17 | Code |
| 8 | X-Pool | 51.2 | Yes | X-Pool: Cross-Modal Language-Video Attention for... | 2022-03-28 | Code |
| 9 | EMCL-Net | 49.2 | No | Expectation-Maximization Contrastive Learning fo... | 2022-11-21 | Code |
| 10 | CLIP | 22.1 | No | A Straightforward Framework For Video Retrieval ... | 2021-02-24 | Code |