Video-to-Text Retrieval on MUGEN n=500
86.6AccuracyDOLPHIN + CLIPPERT
Evaluation Results
| Method | Links | |
|---|---|---|
| DOLPHIN + CLIPPERTbackbone=DOLPHIN, extension=CLIPPERT2026.02 | 86.6 | |
| SCALLOP + CLIPPERTbackbone=SCALLOP, extension=CLIPPERT2026.02 | 74.9 | |
| DOLPHINbackbone=DOLPHIN2026.02 | 33.9 | |
| SCALLOPbackbone=SCALLOP2026.02 | 26.39 |