Compositional Image Synthesis with Inference-Time Scaling
This addresses compositionality issues in text-to-image generation for users needing more accurate and reliable image synthesis from prompts, though it appears incremental as it builds on existing methods like LLMs and VLMs.
The paper tackles the problem of compositionality in text-to-image models, which often fail to render accurate object counts, attributes, and spatial relations, by proposing a training-free framework that combines an object-centric approach with self-refinement to improve layout faithfulness while preserving aesthetic quality. The result is stronger scene alignment with prompts compared to recent text-to-image models.
Despite their impressive realism, modern text-to-image models still struggle with compositionality, often failing to render accurate object counts, attributes, and spatial relations. To address this challenge, we present a training-free framework that combines an object-centric approach with self-refinement to improve layout faithfulness while preserving aesthetic quality. Specifically, we leverage large language models (LLMs) to synthesize explicit layouts from input prompts, and we inject these layouts into the image generation process, where a object-centric vision-language model (VLM) judge reranks multiple candidates to select the most prompt-aligned outcome iteratively. By unifying explicit layout-grounding with self-refine-based inference-time scaling, our framework achieves stronger scene alignment with prompts compared to recent text-to-image models. The code are available at https://github.com/gcl-inha/ReFocus.