AI RESEARCH
Training-Free Object-Background Compositional T2I via Dynamic Spatial Guidance and Multi-Path Pruning
arXiv CS.CV
•
ArXi:2604.09850v1 Announce Type: new Existing text-to-image diffusion models, while excelling at subject synthesis, exhibit a persistent foreground bias that treats the background as a passive and under-optimized byproduct. This imbalance compromises global scene coherence and constrains compositional control. To address the limitation, we propose a