1 paper · 1 filter
Michael Saxon, Yiran Luo, Sharon Levy +3
Benchmarks of the multilingual capabilities of text-to-image (T2I) models compare generated images prompted in a test language to an expected image distribution over a concept set.…