Research map / Image generation / Text-to-image diffusion
HP-Edit: A Human-Preference Post-Training Framework for Image Editing
Fan Li, Chonghuinan Wang, Lina Lei, Yuping Qiu, Jiaqi Xu, Jiaxiu Jiang, Xinran Qin, Zhikai Chen, Fenglong Song, Zhixin Wang, Renjing Pei, Wangmeng Zuo
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| post-train with HP-Edit | Qwen-Image-Edit-2509 · RealPref-Bench | HP-score (overall): 4.472 -> 4.667 | Improved |
| post-train with HP-Edit (relighting task) | Qwen-Image-Edit-2509 · RealPref-Bench (Relighting) | HP-score: 3.54 -> 3.913 | Improved |
| train with unfiltered BaseData and basic scorer (ablation) | Qwen-Image-Edit-2509 · RealPref-Bench | HP-score (overall): 4.472 -> 4.391 | Hurt |
| filter out score-5 cases (RealPref-50K) instead of raw BaseData (ablation) | HP-Edit with Base-Scorer · RealPref-Bench | HP-score (overall): 4.391 -> 4.577 | Improved |
| add task-specific HP-Scorer instead of basic scorer (ablation) | HP-Edit with RealPref-50K · RealPref-Bench | HP-score (overall): 4.577 -> 4.667 | Improved |
| increase LoRA rank from 8 to 32 (ablation) | HP-Edit · RealPref-Bench | HP-score (overall): 4.614 -> 4.667 | Improved |
| increase LoRA rank from 32 to 128 (ablation) | HP-Edit · RealPref-Bench | HP-score (overall): 4.667 -> 4.645 | No clear effect |
| use online GRPO with HP-Scorer instead of offline DPO (ablation) | HP-Edit · RealPref-Bench | HP-score (overall): 4.521 -> 4.590 | Improved |
| apply HP-Edit post-training | Qwen-Image-Edit-2509 · GEdit-Bench-EN | G_O: 7.54 -> 8.30 | Improved |
| apply HP-Edit post-training | Qwen-Image-Edit-2509 · DreamBench++ | CP times PF: 0.575 -> 0.630 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- X2Edit — X2Edit: Revisiting Arbitrary-Instruction Image Editing Through Self-Constructed Data and Task-Aware Representation Learning (AAAI 2026) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- HIVE: Harnessing Human Feedback for Instructional Visual Editing (CVPR 2024 · similarity 0.90)
- Personalized Image Editing in Text-to-Image Diffusion Models via Collaborative Direct Preference Optimization (NeurIPS 2025 · similarity 0.89)
- Multimodal Large Language Models Make Text-to-Image Generative Models Align Better (NeurIPS 2024 · similarity 0.87)
- David and Goliath: Small One-step Model Beats Large Diffusion with Score Post-training (ICML 2025 · similarity 0.87)
- Edit-R2: Context-Aware Reinforcement Learning for Multi-Turn Image Editing (NeurIPS 2026 · similarity 0.87)
- Multi-Reward as Condition for Instruction-based Image Editing (ICLR 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
