Guiding What Not to Generate: Automated Negative Prompting for Text-Image Alignment
This addresses the problem of inaccurate image generation for users of diffusion models, offering an incremental improvement through automated negative prompts.
The paper tackles the challenge of precise text-image alignment in text-to-image generation, particularly for complex prompts, by introducing an automated negative prompting pipeline that improves alignment, achieving scores of 0.571 vs. 0.371 on GenEval++ and best overall performance on Imagine-Bench.
Despite substantial progress in text-to-image generation, achieving precise text-image alignment remains challenging, particularly for prompts with rich compositional structure or imaginative elements. To address this, we introduce Negative Prompting for Image Correction (NPC), an automated pipeline that improves alignment by identifying and applying negative prompts that suppress unintended content. We begin by analyzing cross-attention patterns to explain why both targeted negatives-those directly tied to the prompt's alignment error-and untargeted negatives-tokens unrelated to the prompt but present in the generated image-can enhance alignment. To discover useful negatives, NPC generates candidate prompts using a verifier-captioner-proposer framework and ranks them with a salient text-space score, enabling effective selection without requiring additional image synthesis. On GenEval++ and Imagine-Bench, NPC outperforms strong baselines, achieving 0.571 vs. 0.371 on GenEval++ and the best overall performance on Imagine-Bench. By guiding what not to generate, NPC provides a principled, fully automated route to stronger text-image alignment in diffusion models. Code is released at https://github.com/wiarae/NPC.