Training-Free Text-to-Image Compositional Food Generation via Prompt Grafting
This addresses a domain-specific challenge for applications like image-based dietary assessment and recipe visualization, offering an incremental improvement over existing methods.
The paper tackles the problem of generating accurate multi-food images with text-to-image diffusion models, which often suffer from object entanglement, by introducing Prompt Grafting, a training-free framework that improves target object presence and enables controllable separation.
Real-world meal images often contain multiple food items, making reliable compositional food image generation important for applications such as image-based dietary assessment, where multi-food data augmentation is needed, and recipe visualization. However, modern text-to-image diffusion models struggle to generate accurate multi-food images due to object entanglement, where adjacent foods (e.g., rice and soup) fuse together because many foods do not have clear boundaries. To address this challenge, we introduce Prompt Grafting (PG), a training-free framework that combines explicit spatial cues in text with implicit layout guidance during sampling. PG runs a two-stage process where a layout prompt first establishes distinct regions and the target prompt is grafted once layout formation stabilizes. The framework enables food entanglement control: users can specify which food items should remain separated or be intentionally mixed by editing the arrangement of layouts. Across two food datasets, our method significantly improves the presence of target objects and provides qualitative evidence of controllable separation.