TC-CLIP

Reported on 4 benchmarks across 1 task · 1 paper · 2 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Computer Vision4 results

Zero-Shot Action RecognitiononKinetics
Top-1 Accuracy· 2024-04-15
78.1
SOTA
Leveraging Temporal Contextualization for Video Action Recognition arXiv:2404.09490
Zero-Shot Action RecognitiononKinetics
Top-5 Accuracy· 2024-04-15
95.7
SOTA
Leveraging Temporal Contextualization for Video Action Recognition arXiv:2404.09490
Zero-Shot Action RecognitiononUCF101
Top-1 Accuracy· 2024-04-15
85.4
best: 92.8 (OTI(ViT-L/14))
Leveraging Temporal Contextualization for Video Action Recognition arXiv:2404.09490
Zero-Shot Action RecognitiononHMDB51
Top-1 Accuracy· 2024-04-15
56
best: 64.7 (MOV (ViT-L/14))
Leveraging Temporal Contextualization for Video Action Recognition arXiv:2404.09490