1 paper
Utkarsh Nath, Rajeev Goel, Rahul Khurana +5
Current text-to-3D methods excel at generating single objects but falter on compositional prompts. We argue this failure is fundamental to their optimization schedules, as simultan…