AI RESEARCH
PSR: Scaling Multi-Subject Personalized Image Generation with Pairwise Subject-Consistency Rewards
arXiv CS.CV
•
ArXi:2512.01236v2 Announce Type: replace Personalized generation models for a single subject have nstrated remarkable effectiveness, highlighting their significant potential. However, when extended to multiple subjects, existing models often exhibit degraded performance, particularly in maintaining subject consistency and adhering to textual prompts. We attribute these limitations to the absence of high-quality multi-subject datasets and refined post-