1 paper
Parth Bhalerao, Mounika Yalamarty, Brian Trinh +1
Text-to-image generation models have achieved strong performance in culturally homogeneous settings, yet their ability to generate multicultural scenes, where people and landmarks…