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

VidHalluc

Benchmarks

Task NameDataset NameSOTA ResultTrend
Hallucination examinationVidHalluc
BQA78.08
29
Video UnderstandingVidHalluc
Accuracy83.84
18
Video ReasoningVidHalluc (test)
Binary QA Accuracy (ACH)81.15
13
Hallucination DetectionVidHalluc 19
MC Accuracy91
4
Showing 4 of 4 rows