Image-POSER: Reflective RL for Multi-Expert Image Generation and Editing

TL;DR

Image-POSER introduces a reinforcement learning framework that dynamically orchestrates multiple image generation and editing experts to handle complex, long-form prompts, improving alignment, fidelity, and aesthetics in AI-generated images.

Contribution

The paper presents a novel reflective RL approach that manages diverse pretrained models for complex image synthesis and editing tasks, enabling end-to-end long-form prompt handling.

Findings

Outperforms baseline models in alignment, fidelity, and aesthetics

Achieves higher human preference scores

Demonstrates effective long-form prompt decomposition and expert pipeline learning

Abstract

Recent advances in text-to-image generation have produced strong single-shot models, yet no individual system reliably executes the long, compositional prompts typical of creative workflows. We introduce Image-POSER, a reflective reinforcement learning framework that (i) orchestrates a diverse registry of pretrained text-to-image and image-to-image experts, (ii) handles long-form prompts end-to-end through dynamic task decomposition, and (iii) supervises alignment at each step via structured feedback from a vision-language model critic. By casting image synthesis and editing as a Markov Decision Process, we learn non-trivial expert pipelines that adaptively combine strengths across models. Experiments show that Image-POSER outperforms baselines, including frontier models, across industry-standard and custom benchmarks in alignment, fidelity, and aesthetics, and is consistently preferred in human evaluations. These results highlight that reinforcement learning can endow AI systems with the capacity to autonomously decompose, reorder, and combine visual models, moving towards general-purpose visual assistants.