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

T2I-FactualBench-SKCM

Benchmarks

Task NameDataset NameSOTA ResultTrend
Text-to-Image Factuality EvaluationT2I-FactualBench-SKCM 100
FAGER88.93
6
Factual A/B testT2I-FactualBench-SKCM 100 pairs (test)
Pairwise Accuracy87
3
Showing 2 of 2 rows