Enhancing Spatial Understanding in Image Generation via Reward Modeling
This work addresses the problem of complex spatial encoding in text-to-image generation for users needing precise visual outputs, representing an incremental improvement in a specific domain.
The paper tackles the challenge of generating images with accurate spatial relationships from text prompts by introducing a reward model trained on a dataset of 80k preference pairs, which improves spatial understanding and outperforms proprietary models in evaluations.
Recent progress in text-to-image generation has greatly advanced visual fidelity and creativity, but it has also imposed higher demands on prompt complexity-particularly in encoding intricate spatial relationships. In such cases, achieving satisfactory results often requires multiple sampling attempts. To address this challenge, we introduce a novel method that strengthens the spatial understanding of current image generation models. We first construct the SpatialReward-Dataset with over 80k preference pairs. Building on this dataset, we build SpatialScore, a reward model designed to evaluate the accuracy of spatial relationships in text-to-image generation, achieving performance that even surpasses leading proprietary models on spatial evaluation. We further demonstrate that this reward model effectively enables online reinforcement learning for the complex spatial generation. Extensive experiments across multiple benchmarks show that our specialized reward model yields significant and consistent gains in spatial understanding for image generation.