Multimodal LLM-Guided Semantic Correction in Text-to-Image Diffusion

TL;DR

This paper introduces PPAD, a novel framework that uses a Multimodal Large Language Model to analyze and correct semantic errors during the diffusion process in text-to-image generation, improving alignment and quality.

Contribution

It is the first to incorporate real-time semantic supervision via MLLM during diffusion, enabling active correction with minimal steps and enhancing image fidelity.

Findings

Significant improvement in prompt-image alignment.

Effective semantic correction with few diffusion steps.

Versatile framework supporting inference-only and training modes.

Abstract

Diffusion models have become the mainstream architecture for text-to-image generation, achieving remarkable progress in visual quality and prompt controllability. However, current inference pipelines generally lack interpretable semantic supervision and correction mechanisms throughout the denoising process. Most existing approaches rely solely on post-hoc scoring of the final image, prompt filtering, or heuristic resampling strategies-making them ineffective in providing actionable guidance for correcting the generative trajectory. As a result, models often suffer from object confusion, spatial errors, inaccurate counts, and missing semantic elements, severely compromising prompt-image alignment and image quality. To tackle these challenges, we propose MLLM Semantic-Corrected Ping-Pong-Ahead Diffusion (PPAD), a novel framework that, for the first time, introduces a Multimodal Large Language Model (MLLM) as a semantic observer during inference. PPAD performs real-time analysis on intermediate generations, identifies latent semantic inconsistencies, and translates feedback into controllable signals that actively guide the remaining denoising steps. The framework supports both inference-only and training-enhanced settings, and performs semantic correction at only extremely few diffusion steps, offering strong generality and scalability. Extensive experiments demonstrate PPAD's significant improvements.