HP-Edit: A Human-Preference Post-Training Framework for Image Editing

Hugging Face Daily Papers Papers
image-editing rlhf diffusion-models human-preferences post-training

Summary

HP-Edit introduces a post-training framework that aligns diffusion-based image editing models with human preferences via RLHF, using a new 50K real-world dataset and an automatic VLM-based evaluator.

Common image editing tasks typically adopt powerful generative diffusion models as the leading paradigm for real-world content editing. Meanwhile, although reinforcement learning (RL) methods such as Diffusion-DPO and Flow-GRPO have further improved generation quality, efficiently applying Reinforcement Learning from Human Feedback (RLHF) to diffusion-based editing remains largely unexplored, due to a lack of scalable human-preference datasets and frameworks tailored to diverse editing needs. To fill this gap, we propose HP-Edit, a post-training framework for Human Preference-aligned Editing, and introduce RealPref-50K, a real-world dataset across eight common tasks and balancing common object editing. Specifically, HP-Edit leverages a small amount of human-preference scoring data and a pretrained visual large language model (VLM) to develop HP-Scorer--an automatic, human preference-aligned evaluator. We then use HP-Scorer both to efficiently build a scalable preference dataset and to serve as the reward function for post-training the editing model. We also introduce RealPref-Bench, a benchmark for evaluating real-world editing performance. Extensive experiments demonstrate that our approach significantly enhances models such as Qwen-Image-Edit-2509, aligning their outputs more closely with human preference.
Original Article
View Cached Full Text

Cached at: 04/22/26, 10:35 AM

Paper page - HP-Edit: A Human-Preference Post-Training Framework for Image Editing

Source: https://huggingface.co/papers/2604.19406 Authors:

,

,

,

,

,

,

,

,

,

,

Abstract

A post-training framework called HP-Edit is introduced to align image editing models with human preferences using a novel automatic evaluator and a real-world dataset, improving editing quality through reinforcement learning techniques.

Commonimage editingtasks typically adopt powerfulgenerative diffusion modelsas the leading paradigm for real-world content editing. Meanwhile, althoughreinforcement learning(RL) methods such as Diffusion-DPO and Flow-GRPO have further improved generation quality, efficiently applyingReinforcement Learningfrom Human Feedback (RLHF) to diffusion-based editing remains largely unexplored, due to a lack of scalable human-preference datasets and frameworks tailored to diverse editing needs. To fill this gap, we propose HP-Edit, apost-trainingframework forHuman Preference-aligned Editing, and introduce RealPref-50K, a real-world dataset across eight common tasks and balancing common object editing. Specifically, HP-Edit leverages a small amount of human-preference scoring data and a pretrainedvisual large language model(VLM) to developHP-Scorer--an automatic, human preference-aligned evaluator. We then useHP-Scorerboth to efficiently build a scalablepreference datasetand to serve as the reward function forpost-trainingthe editing model. We also introduce RealPref-Bench, a benchmark for evaluating real-world editing performance. Extensive experiments demonstrate that our approach significantly enhances models such as Qwen-Image-Edit-2509, aligning their outputs more closely with human preference.

View arXiv pageView PDFAdd to collection

Get this paper in your agent:

hf papers read 2604\.19406

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2604.19406 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2604.19406 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2604.19406 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

Uni-Edit: Intelligent Editing Is A General Task For Unified Model Tuning

Hugging Face Daily Papers

Uni-Edit proposes using intelligent image editing as a single general task to simultaneously improve unified multimodal models' understanding, generation, and editing capabilities, with an automated data synthesis pipeline creating complex editing instructions.

Qwen-Image-2.0-RL Technical Report

Hugging Face Daily Papers

This technical report presents Qwen-Image-2.0-RL, a post-training pipeline using reinforcement learning from human feedback and on-policy distillation to enhance visual quality and instruction-following in image generation and editing tasks.