Fetching the paper…

Be Yourself: Bounded Attention for Multi-Subject Text-to-Image Generation · Around