Zero-shot spatial layout conditioning for text-to-image diffusion models

TL;DR

This paper introduces ZestGuide, a zero-shot segmentation guidance method for text-to-image diffusion models that enables precise spatial control using implicit segmentation maps without additional training.

Contribution

It presents a novel zero-shot approach that integrates segmentation guidance into pre-trained diffusion models, improving spatial accuracy without extra training.

Findings

Enhanced spatial alignment with input masks

Improved mIoU scores on COCO dataset

Maintained high image quality with better segmentation accuracy

Abstract

Large-scale text-to-image diffusion models have significantly improved the state of the art in generative image modelling and allow for an intuitive and powerful user interface to drive the image generation process. Expressing spatial constraints, e.g. to position specific objects in particular locations, is cumbersome using text; and current text-based image generation models are not able to accurately follow such instructions. In this paper we consider image generation from text associated with segments on the image canvas, which combines an intuitive natural language interface with precise spatial control over the generated content. We propose ZestGuide, a zero-shot segmentation guidance approach that can be plugged into pre-trained text-to-image diffusion models, and does not require any additional training. It leverages implicit segmentation maps that can be extracted from cross-attention layers, and uses them to align the generation with input masks. Our experimental results combine high image quality with accurate alignment of generated content with input segmentations, and improve over prior work both quantitatively and qualitatively, including methods that require training on images with corresponding segmentations. Compared to Paint with Words, the previous state-of-the art in image generation with zero-shot segmentation conditioning, we improve by 5 to 10 mIoU points on the COCO dataset with similar FID scores.