Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

WikiVideo

Benchmarks

Task NameDataset NameSOTA ResultTrend
Article GenerationWikiVideo (test)
InfoP Score94.5
10
Multimodal RetrievalWikiVideo (test)
Alpha-nDCG62.8
10
UMUI Judgment CalibrationWIKIVIDEO (v)
MSE0.0784
8
Scalar probability judgmentWikiVideo Audio-only
MSE (x100)3.4
5
Scalar probability judgmentWikiVideo Vision-only
MSE0.078
5
Binary JudgmentWikiVideo Audio-only
Accuracy71.5
5
Binary JudgmentWikiVideo Vision-only
Accuracy81.1
5
UMUI Judgment CalibrationWIKIVIDEO (A)
MSE0.0335
5
Scalar probability judgmentWikiVideo Audio-Visual
MSE (x100)7.9
3
Binary JudgmentWikiVideo Audio-Visual
Accuracy70.1
3
Showing 10 of 10 rows