RAISE: Requirement-Adaptive Evolutionary Refinement for Training-Free Text-to-Image Alignment

TL;DR

RAISE introduces a training-free, adaptive evolutionary framework that improves text-to-image alignment by dynamically refining images at inference time based on requirement satisfaction, reducing computation and enhancing fidelity.

Contribution

It presents a novel requirement-driven evolutionary method for inference-time image refinement that adapts to prompt complexity without additional training or fine-tuning.

Findings

Achieves state-of-the-art alignment scores on GenEval with fewer samples.

Reduces generated samples by 30-40% and VLM calls by 80% compared to prior methods.

Demonstrates effective, model-agnostic self-improvement across benchmarks.

Abstract

Recent text-to-image (T2I) diffusion models achieve remarkable realism, yet faithful prompt-image alignment remains challenging, particularly for complex prompts with multiple objects, relations, and fine-grained attributes. Existing training-free inference-time scaling methods rely on fixed iteration budgets that cannot adapt to prompt difficulty, while reflection-tuned models require carefully curated reflection datasets and extensive joint fine-tuning of diffusion and vision-language models, often overfitting to reflection paths data and lacking transferability across models. We introduce RAISE (Requirement-Adaptive Self-Improving Evolution), a training-free, requirement-driven evolutionary framework for adaptive T2I generation. RAISE formulates image generation as a requirement-driven adaptive scaling process, evolving a population of candidates at inference time through a diverse set of refinement actions-including prompt rewriting, noise resampling, and instructional editing. Each generation is verified against a structured checklist of requirements, enabling the system to dynamically identify unsatisfied items and allocate further computation only where needed. This achieves adaptive test-time scaling that aligns computational effort with semantic query complexity. On GenEval and DrawBench, RAISE attains state-of-the-art alignment (0.94 overall GenEval) while incurring fewer generated samples (reduced by 30-40%) and VLM calls (reduced by 80%) than prior scaling and reflection-tuned baselines, demonstrating efficient, generalizable, and model-agnostic multi-round self-improvement. Code is available at https://github.com/LiyaoJiang1998/RAISE.