Evaluation-Verification Reward for Consistent Multi-Reference Image Editing

Hugging Face Daily Papers Papers
image-editing multi-reference reinforcement-learning reward-model mllm consistency

Summary

This paper introduces a Multi-dimensional Evaluation-Verification Reward (EVR) for reinforcement learning fine-tuning of multi-reference image editing models, improving visual consistency and harmony.

While recent image editing models have made rapid progress, multi-reference editing remains challenging, particularly in maintaining visual consistency across references and ensuring overall visual harmony. Reinforcement learning has proven highly effective for text-to-image generation and single-image editing, but its extension to multi-reference editing is hindered by the absence of suitable reward models that capture multi-image relational constraints. Moreover, naively using multimodal large language models(MLLMs) as zero-shot evaluators faces a key tension between hallucination-prone long-form reasoning and the limited deductive power of short-form judgments. We address these issues with a Multi-dimensional Evaluation-Verification Reward(EVR). EVR decomposes evaluation into distinct visual criteria; for each criterion, an MLLM Evaluator generates multiple candidate hypotheses, and a Verifier grounds each claim in concrete visual evidence to accept or reject it, producing reliable and fine-grained reward signals. Together with a scalable data pipeline, our method enables RL fine-tuning of off-the-shelf editors without architectural changes. Extensive experiments show substantial gains over the base Qwen-Image-Edit, improving consistency and harmony to match or surpass NanoBanana.
Original Article
View Cached Full Text

Cached at: 08/03/26, 05:30 AM

Paper page - Evaluation-Verification Reward for Consistent Multi-Reference Image Editing

Source: https://huggingface.co/papers/2607.29025

Abstract

Whilerecentimageeditingmodelshavemaderapidprogress,multi-referenceeditingremainschallenging,particularlyinmaintainingvisualconsistencyacrossreferencesandensuringoverallvisualharmony.Reinforcementlearninghasprovenhighlyeffectivefortext-to-imagegenerationandsingle-imageediting,butitsextensiontomulti-referenceeditingishinderedbytheabsenceofsuitablerewardmodelsthatcapturemulti-imagerelationalconstraints.Moreover,naivelyusingmultimodallargelanguagemodels(MLLMs)aszero-shotevaluatorsfacesakeytensionbetweenhallucination-pronelong-formreasoningandthelimiteddeductivepowerofshort-formjudgments.WeaddresstheseissueswithaMulti-dimensionalEvaluation-VerificationReward(EVR).EVRdecomposesevaluationintodistinctvisualcriteria;foreachcriterion,anMLLMEvaluatorgeneratesmultiplecandidatehypotheses,andaVerifiergroundseachclaiminconcretevisualevidencetoacceptorrejectit,producingreliableandfine-grainedrewardsignals.Togetherwithascalabledatapipeline,ourmethodenablesRLfine-tuningofoff-the-shelfeditorswithoutarchitecturalchanges.ExtensiveexperimentsshowsubstantialgainsoverthebaseQwen-Image-Edit,improvingconsistencyandharmonytomatchorsurpassNanoBanana.

View arXiv pageView PDFAdd to collection

Get this paper in your agent:

hf papers read 2607\.29025

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/2607.29025 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2607.29025 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2607.29025 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

Is This Edit Correct? A Multi-Dimensional Benchmark for Reasoning-Aware Image Editing

Hugging Face Daily Papers

This paper introduces RE-Edit, a benchmark for evaluating image editing systems across five reasoning dimensions (physical, environmental, cultural, causal, referential) to assess logical consistency beyond visual plausibility. The benchmark includes 1,000 samples and evaluates ten open-source and two commercial models, showing that even advanced systems struggle with implicit multi-dimensional reasoning.

Video Models Can Reason with Verifiable Rewards

Hugging Face Daily Papers

VideoRLVR optimizes video diffusion models for verifiable reasoning tasks using reinforcement learning with rule-based rewards, achieving better performance than supervised methods in constraint-satisfying video generation.