RealGen: Photorealistic Text-to-Image Generation via Detector-Guided Rewards

2025-11-29

AI Summary

Current text-to-image models (e.g., GPT-Image-1, Qwen-Image) suffer from prominent AI artifacts—such as oversmoothed skin and unrealistic facial specular highlights—hindering photorealistic fidelity. To address this, we propose Detector Reward, a novel reward mechanism that jointly quantifies semantic and feature-level artifacts via multi-granularity image detectors, enabling end-to-end reward-driven optimization. Our framework unifies LLM-based prompt refinement, diffusion-based image generation, and GRPO-based reinforcement learning for joint training. Furthermore, we introduce RealBench—the first automated benchmark for photorealism evaluation—designed to assess perceptual realism, detail fidelity, and aesthetic quality. Extensive experiments demonstrate that our method consistently outperforms GPT-Image-1, Qwen-Image, and FLUX-Krea across all dimensions. Notably, RealBench scores exhibit strong correlation with human perceptual judgments, validating its effectiveness as an objective realism metric.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Generation

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

With the continuous advancement of image generation technology, advanced models such as GPT-Image-1 and Qwen-Image have achieved remarkable text-to-image consistency and world knowledge However, these models still fall short in photorealistic image generation. Even on simple T2I tasks, they tend to produce " fake" images with distinct AI artifacts, often characterized by "overly smooth skin" and "oily facial sheens". To recapture the original goal of "indistinguishable-from-reality" generation, we propose RealGen, a photorealistic text-to-image framework. RealGen integrates an LLM component for prompt optimization and a diffusion model for realistic image generation. Inspired by adversarial generation, RealGen introduces a "Detector Reward" mechanism, which quantifies artifacts and assesses realism using both semantic-level and feature-level synthetic image detectors. We leverage this reward signal with the GRPO algorithm to optimize the entire generation pipeline, significantly enhancing image realism and detail. Furthermore, we propose RealBench, an automated evaluation benchmark employing Detector-Scoring and Arena-Scoring. It enables human-free photorealism assessment, yielding results that are more accurate and aligned with real user experience. Experiments demonstrate that RealGen significantly outperforms general models like GPT-Image-1 and Qwen-Image, as well as specialized photorealistic models like FLUX-Krea, in terms of realism, detail, and aesthetics. The code is available at https://github.com/yejy53/RealGen.

Problem

• Addresses AI artifacts in text-to-image generation
• Enhances photorealism using detector-guided reward mechanisms
• Proposes automated benchmark for realism evaluation

Innovation

• Integrates LLM for prompt optimization and diffusion model
• Introduces Detector Reward mechanism for artifact quantification
• Uses GRPO algorithm to optimize generation pipeline realism