Video Recognition on GYM
90.8AccuracySMILE
Evaluation Results
| Method | Links | |
|---|---|---|
| SMILEProtocol=Finetuning, Pretraining Video Dataset=K4002025.04 | 90.8 | |
| UMTProtocol=Finetuning, Pretraining Video Dataset=K7002025.04 | 89.9 | |
| ViCLIPProtocol=Finetuning, Pretraining Video Dataset=Intervid-10M2025.04 | 89.7 | |
| CLIPProtocol=Finetuning, Pretraining Video Dataset=N/A2025.04 | 88 | |
| SMILEProtocol=Linear Probing, Pretraining Video Dataset=K4002025.04 | 30.2 | |
| ViCLIPProtocol=Linear Probing, Pretraining Video Dataset=Intervid-10M2025.04 | 27.3 | |
| UMTProtocol=Linear Probing, Pretraining Video Dataset=K7002025.04 | 26.4 | |
| CLIPProtocol=Linear Probing, Pretraining Video Dataset=N/A2025.04 | 20.7 |