Action Recognition on Kinetics-700 (test)
84.8AccuracyNS ENet-L2
Evaluation Results
| Method | Links | |
|---|---|---|
| NS ENet-L2Evaluation Protocol=Finetune2021.02 | 84.8 | |
| CoCa-gTraining Data=I-3B, Frames x Resolution=16 x 576, Probing Type=Attentive Probing2024.03 | 81.1 | |
| InternVideo2s2-6BTraining Data=IV-400M, Frames x Resolution=16 x 224, Probing Type=Attentive Probing2024.03 | 81 | |
| InternVideo2s2-1BTraining Data=IV-25.5M, Frames x Resolution=16 x 224, Probing Type=Attentive Probing2024.03 | 79.5 | |
| CLIPEvaluation Protocol=Linear2021.02 | 73 | |
| Baseline I3DEvaluation Protocol=Finetune2021.02 | 70.2 | |
| CLIPEvaluation Protocol=Zero-Shot2021.02 | 69.6 | |
| NS ENet-L2Evaluation Protocol=Linear2021.02 | 68.2 | |
| cf_LArchitecture=SS Conformer, Model Scale=Large, Pre-training=Self-supervised (SS), Evaluation Protocol=Fine-tuning, Training Samples=504,4432021.10 | 23.5 | |
| cf_SArchitecture=SS Conformer, Model Scale=Small, Pre-training=Self-supervised (SS), Evaluation Protocol=Fine-tuning, Training Samples=504,4432021.10 | 20.4 | |
| Prior work [11]Pre-training=Supervised (S), Evaluation Protocol=Shallow classifier, Training Samples=504,4432021.10 | 18 |