CoCa (frozen)

Reported on 4 benchmarks across 1 task · 1 paper

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

Computer Vision4 results

VideoonKinetics-700
Top-1 Accuracy· uses extra data· 2022-05-04
81.1
best: 85.9 (InternVideo2-6B)
CoCa: Contrastive Captioners are Image-Text Foundation Models arXiv:2205.01917
VideoonMoments in Time
Top 1 Accuracy· uses extra data· 2022-05-04
47.4
best: 53.1 (OmniVec2)
CoCa: Contrastive Captioners are Image-Text Foundation Models arXiv:2205.01917
VideoonKinetics-400
Acc@1· 2022-05-04
88
best: 93.6 (OmniVec2)
CoCa: Contrastive Captioners are Image-Text Foundation Models arXiv:2205.01917
VideoonKinetics-600
Top-1 Accuracy· uses extra data· 2022-05-04
88.5
best: 91.9 (InternVideo2-6B)
CoCa: Contrastive Captioners are Image-Text Foundation Models arXiv:2205.01917