Towards Understanding Cross and Self-Attention in Stable Diffusion for Text-Guided Image Editing

arXiv:2403.03431v1 Announce Type: new
Abstract: Deep Text-to-Image Synthesis (TIS) models such as Stable Diffusion have recently gained significant popularity for creative Text-to-image generation. Yet, for domain-specific scenarios, tuning-free Text-guided Image Editing (TIE) is of greater importance for application developers, which modify objects or object properties in images by manipulating feature components in attention layers during the generation process. However, little is known about what semantic meanings these attention layers have learned and which parts of the attention maps contribute to …

abstract application arxiv attention creative cs.cv developers diffusion domain editing free image image generation images importance object objects self-attention stable diffusion synthesis text text-to-image type understanding