LocRef-Diffusion:Tuning-Free Layout and Appearance-Guided Generation

TL;DR

LocRef-Diffusion is a tuning-free diffusion model enabling personalized, controllable image generation with precise instance placement and appearance matching, advancing the capabilities of text-to-image synthesis.

Contribution

The paper introduces LocRef-Diffusion, a novel tuning-free model with layout and appearance control, utilizing Layout-net and appearance-net modules for enhanced personalization.

Findings

Achieves state-of-the-art results on COCO and OpenImages datasets.

Effectively controls instance layout and appearance in generated images.

Demonstrates superior performance over existing methods.

Abstract

Recently, text-to-image models based on diffusion have achieved remarkable success in generating high-quality images. However, the challenge of personalized, controllable generation of instances within these images remains an area in need of further development. In this paper, we present LocRef-Diffusion, a novel, tuning-free model capable of personalized customization of multiple instances' appearance and position within an image. To enhance the precision of instance placement, we introduce a Layout-net, which controls instance generation locations by leveraging both explicit instance layout information and an instance region cross-attention module. To improve the appearance fidelity to reference images, we employ an appearance-net that extracts instance appearance features and integrates them into the diffusion model through cross-attention mechanisms. We conducted extensive experiments on the COCO and OpenImages datasets, and the results demonstrate that our proposed method achieves state-of-the-art performance in layout and appearance guided generation.