Text-to-image (T2I) diffusion models generate high-quality images but often fail to capture the spatial relations specified in text prompts. This limitation can be traced to two factors: lack of fine-grained spatial supervision in training data and inability of text embeddings to encode spatial semantics. We introduce InfSplign, a training-free inference-time method that improves spatial alignment by adjusting the noise through a compound loss in every denoising step. Proposed loss leverages different levels of cross-attention maps extracted from the backbone decoder to enforce accurate object placement and a balanced object presence during sampling. The method is lightweight, plug-and-play, and compatible with any diffusion backbone. Our comprehensive evaluations on VISOR and T2I-CompBench show that InfSplign establishes a new state-of-the-art (to the best of our knowledge), achieving substantial performance gains over the strongest existing inference-time baselines and even outperforming the fine-tuning-based methods. Codebase is available at GitHub.
@article{arxiv.2512.17851,
title = {InfSplign: Inference-Time Spatial Alignment of Text-to-Image Diffusion Models},
author = {Sarah Rastegar and Violeta Chatalbasheva and Sieger Falkena and Anuj Singh and Yanbo Wang and Tejas Gokhale and Hamid Palangi and Hadi Jamali-Rad},
journal= {arXiv preprint arXiv:2512.17851},
year = {2025}
}