Video Retrieval
The objective of video retrieval is as follows: given a text query and a pool of candidate videos, select the video which corresponds to the text query. Typically, the videos are returned as a ranked list of candidates and scored via document retrieval metrics.
Papers
Showing 1–10 of 486 papers
All datasetsMSR-VTT-1kADiDeMoMSR-VTTLSMDCActivityNetMSVDYouCook2FIVR-200KVATEXQuerYDSSv2-label retrievalSSv2-template retrieval
Benchmark Results
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | Aurora (ours, r=64) | text-to-video R@5 | 77.4 | — | Unverified |
| 2 | InternVideo2-6B | text-to-video R@1 | 74.2 | — | Unverified |
| 3 | vid-TLDR (UMT-L) | text-to-video R@1 | 72.3 | — | Unverified |
| 4 | VAST | text-to-video R@1 | 72 | — | Unverified |
| 5 | COSA | text-to-video R@1 | 70.5 | — | Unverified |
| 6 | UMT-L (ViT-L/16) | text-to-video R@1 | 70.4 | — | Unverified |
| 7 | GRAM | text-to-video R@1 | 67.3 | — | Unverified |
| 8 | VALOR | text-to-video R@1 | 61.5 | — | Unverified |
| 9 | TESTA (ViT-B/16) | text-to-video R@1 | 61.2 | — | Unverified |
| 10 | VindLU | text-to-video R@1 | 61.2 | — | Unverified |