Video-to-Text Retrieval on MUGEN n=250
83.7AccuracyDOLPHIN + CLIPPERT
Evaluation Results
| Method | Links | |
|---|---|---|
| DOLPHIN + CLIPPERTbackbone=DOLPHIN, extension=CLIPPERT2026.02 | 83.7 | |
| DOLPHINbackbone=DOLPHIN2026.02 | 33.8 | |
| SCALLOP + CLIPPERTbackbone=SCALLOP, extension=CLIPPERT2026.02 | 27.1 | |
| SCALLOPbackbone=SCALLOP2026.02 | 25.2 |