PrefLUT: Reusable and Refinable Personalized Color Editing from Pairwise Preferences

Chuanzhi Xu; Langyi Chen; Chengkun Yue; Xuanhua Yin; Boyu Wei; Qingwen Zeng; Zihan Deng; Weidong Cai

cs.CVSep 28, 2026

PrefLUT: Reusable and Refinable Personalized Color Editing from Pairwise Preferences

Authors: Chuanzhi Xu, Langyi Chen, Chengkun Yue, Xuanhua Yin, Boyu Wei, Qingwen Zeng, Zihan Deng, Weidong Cai

Organizations: The University of Sydney · Indiana University · The University of Hong Kong

Abstract

Photographic color editing is inherently personal: the same image can appear too warm, too muted, or already satisfactory to different users. Most lookup table (LUT) and reference-guided methods target a specified appearance rather than model persistent preferences from repeated user choices. To address this gap, we introduce PrefLUT, a reusable and refinable user-preference modeling framework for deployable 3D LUTs, encoding ordered preferred/non-preferred image pairs into a lightweight Reusable User Profile that is reused across queries and refined using additional user preference pairs, without per-user optimization. A Query-Conditioned LUT Predictor combines this profile with each image to predict a LUT latent vector and edit strength. An Identity-Residual LUT Decoder and Edit-Strength Controller then produce an exportable 3D LUT. Experiments on three datasets demonstrate effective personalized editing and general-purpose enhancement. Each quantized profile requires only 260 bytes, and editing takes 1.365 ms/image on an RTX 5090 GPU. We also introduce the Preference-Conditioning Verification Protocol (PCVP), an evaluation protocol to verify whether personalized image edits depend on user preferences and the query image through controlled changes to user profiles, preference orders, pair correspondences, and query images.

Figures & tables

Figure 1: Preference-conditioned editing across users. Examples from two PPSD users show ordered preferred/non-preferred image pairs (top) and unseen inputs with PrefLUT outputs (bottom). PrefLUT outputs reflect each user’s color preferences across different scenes.

Figure 2: PrefLUT pipeline. Profile Construction (top) encodes ordered preference image pairs into a Reusable User Profile that can be refined with new feedback. Query-Time Editing (bottom) combines the stored profile with each query image to predict a 3D LUT for full-resolution editing.

	(A) Direct Fidelity to Preferred				(B) Personalized Editing Quality				(C) Model Size and Inference Efficiency
	Output vs. Target				CQS $\uparrow$				Params	Compute	Profile	Construct.	Editing
Method	$\Delta E_{00}\downarrow$	LPIPS $\downarrow$	PSNR $\uparrow$	SSIM $\uparrow$	$\Delta E_{00}\uparrow$	LPIPS $\uparrow$	PSNR $\uparrow$	SSIM $\uparrow$	(M)	(GFLOPs)	(B/user)	(ms/user)	(ms/image)
PieNet †	8.432	0.118	21.719	0.818	0.218	14.978	36.611	0.901	26.349	76.094	2,048	130.626	4.091
StarEnhancer †	8.077	0.090	21.949	0.835	0.213	21.477	30.960	0.918	38.052	3.638	4,096	141.096	3.253
PIE-MSM †	8.179	0.092	21.576	0.829	0.173	17.711	26.827	0.905	90.756	55.909	32,768	119.766	9.401
DiffRetouch †	9.380	0.108	20.881	0.823	0.121	13.184	22.328	0.884	924.139	15,939.293	16	12.705	653.309

Table 1: Personalized editing quality and efficiency on PPSD. (A) Direct fidelity to preferred target images (Preferred) for edited non-preferred queries. (B) Personalized editing quality measured by CQS. (C) Model size and inference efficiency.

\dagger

marks baselines adapted to PPSD.

Figure 3: Qualitative results on PPSD. Input is the non-preferred query and Preferred its paired target. (A) Baseline comparisons. (B) Correct (the intended user’s profile and current query image) versus five PCVP controls. (C) Profile refinement from

N=4

24

feedback pairs with frozen network weights. Overlays report LPIPS/PSNR/SSIM against Preferred in (A,B) (

\downarrow/\uparrow/\uparrow

), and per-output LPIPS-CQS/PSNR-CQS/SSIM-CQS in (C) (all

\uparrow

). More examples are in Appendix H .

Method	Metric	Wrong user		Reversed order		Mismatched pairs		Training mean		Wrong query		Full Pass
PieNet † (ECCV 2020)	$\Delta E_{00}$	-0.00006		+0.00001		+0.00068		-0.00360		+0.00454
	LPIPS	-0.07606		+0.07724		+0.11300		-0.33478		-0.35530
	PSNR	-0.10056		+0.14574		+0.22164		-0.64456		-0.21505
	SSIM	-0.00059	0/4	-0.00027	0/4	+0.00035	0/4	-0.00169	0/4	+0.00011	1/4	No (0/5)
StarEnhancer † (ICCV 2021)	$\Delta E_{00}$	+0.00145		+0.00334		—		-0.00414		+0.02791
	LPIPS	+0.20574		+0.19325		—		-0.14146		+3.75054

Table 2: PCVP gains (

\Delta\mathrm{CQS}

) are correct-condition CQS minus control-condition CQS. Each test lists gains (left) and the number of passing metrics out of four (right). Bold blue marks passing metrics. A test requires 4/4, and full PCVP requires all five tests. A dash denotes a structurally inapplicable control.

\dagger

marks PPSD-adapted baselines. Evaluation settings are in Appendix D.2 .

	CQS difference				Wrong-user PCVP gain difference
Variant	$\Delta E_{00}$	LPIPS	PSNR	SSIM	$\Delta E_{00}$	LPIPS	PSNR	SSIM
$-$ Explicit Feature Difference	$-0.001062$	$-0.200$	$-0.214$	$-0.000553$	$-0.003074$	$-0.32296$	$-0.28613$	$-0.001116$
$-$ Set Transformer	$-0.019845$	$-1.484$	$-2.250$	$-0.002619$	$-0.004238$	$-0.42888$	$-0.39226$	$-0.001401$
$-$ Query Feature	$-0.003771$	$-1.438$	$+1.482$	$-0.003514$	$-0.003290$	$-0.33886$	$-0.34294$	$-0.001110$
$-$ Wrong-User Contrast Loss	$-0.00000679$	$-0.003$	$+0.003$	$-0.00000439$	$-0.00003627$	$-0.00487$	$-0.00223$	$-0.00001351$

Table 3: Core ablations on PPSD. Differences in CQS and Fixed Wrong-User Profile PCVP gain are reported as variant minus its matched full model. In row order, we remove the Explicit Feature Difference

a_{i}^{+}-a_{i}^{-}

, replace the Set Transformer with masked mean pooling in the Preference Set Aggregator, remove the Query Feature

c_{q}

, and remove the Wrong-User Contrast Loss

\mathcal{L}_{\mathrm{user}}

. All ablations support the contribution of the corresponding components.

Appendix figures & tables9 assets

Supplementary material from the paper’s appendix.

Appendix

PCVP Test	$\Delta E_{00}\uparrow$	LPIPS $\uparrow$	PSNR $\uparrow$	SSIM $\uparrow$
Fixed Wrong-User Profile	$0.003999$ $[\!.002139,\!.006048]$	$0.412528$ $[\!.194432,\!.698803]$	$0.267460$ $[\!.036182,\!.512614]$	$0.001102$ $[\!.000600,\!.001662]$
Reversed-Order Profile	$0.006449$ $[\!.003358,\!.010141]$	$0.383434$ $[\!.109178,\!.764762]$	$0.612272$ $[\!.265722,\!.991219]$	$0.001116$ $[\!.000539,\!.001746]$
Cyclic Wrong-Query Control	$0.015708$ $[\!.012523,\!.019271]$	$2.393584$ $[\!1.629063,\!3.454507]$	$0.881433$ $[\!.773038,\!.990171]$	$0.005383$ $[\!.004403,\!.006354]$

Appendix

Table E1: PrefLUT PCVP gains across evaluation episodes on PPSD. Gains are correct-condition CQS minus the named control’s CQS, combined over sampling seeds 2029–2031 with the model and users fixed. Brackets give paired 95% user-bootstrap confidence intervals.

(A) Unseen-Expert Personalization, 100 Reference Pairs
Method	PSNR $\uparrow$	SSIM $\uparrow$	$\Delta E_{ab}$ $\downarrow$
SpliNet (TIP 2020)	19.94	0.840	13.87
PieNet (ECCV 2020)	20.54	0.851	13.56
StarEnhancer (ICCV 2021)	19.68	0.833	14.94
PIE-MSM (TCSVT 2024)	23.18	0.898	10.13

Appendix

Table E2: MIT-Adobe FiveK results under different evaluation protocols. (A) Published unseen-expert personalization results averaged over Experts A/B/C/D/E. (B) PrefLUT trained for Expert C and evaluated on held-out images, reported as mean

\pm

sample standard deviation over 8 training seeds.

(A) Unseen-Expert Personalization, 100 Reference Pairs
Method	PSNR $\uparrow$	SSIM $\uparrow$	$\Delta E_{ab}$ $\downarrow$
SpliNet (TIP 2020)	20.34	0.865	12.84
PieNet (ECCV 2020)	19.97	0.824	14.32
StarEnhancer (ICCV 2021)	21.22	0.885	12.29
PIE-MSM (TCSVT 2024)	22.97	0.921	9.83

Appendix

Table E3: PPR10K results under different evaluation protocols. (A) Published unseen-expert personalization results averaged over Experts A/B/C. (B) One PrefLUT model trained jointly for these experts, with each row averaged over 2 training seeds. HC denotes human-centered metrics.

	CQS Gain $\uparrow$
Feedback Pairs $N$	$\Delta E_{00}$	LPIPS	PSNR	SSIM
$8$	$+0.002274$ $[0.001578,0.003031]$	$+0.125910$ $[0.073599,0.199634]$	$+0.289340$ $[0.196488,0.379035]$	$+0.000223$ $[0.000117,0.000339]$
$12$	$+0.003037$ $[0.002116,0.004013]$	$+0.172396$ $[0.104210,0.267251]$	$+0.393652$ $[0.255991,0.521696]$	$+0.000350$ $[0.000222,0.000488]$
$16$	$+0.003215$ $[0.002309,0.004167]$	$+0.176357$ $[0.107124,0.271374]$	$+0.443595$ $[0.303425,0.575171]$	$+0.000359$ $[0.000224,0.000503]$
$24$	$+0.003531$ $[0.002684,0.004431]$	$+0.184820$ $[0.114342,0.282361]$	$+0.467157$ $[0.328338,0.597955]$	$+0.000374$ $[0.000237,0.000520]$
$16^{\dagger}$	$+0.003572$ $[0.002074,0.005299]$	$+0.179112$ $[0.083788,0.314092]$	$+0.749985$ $[0.468387,1.039789]$	$+0.000482$ $[0.000263,0.000727]$

Appendix

Table F1: Profile refinement on PPSD.

N

counts accumulated feedback pairs. Gains subtract Frozen Initial Profile CQS (4 pairs) from Cumulative Feedback CQS.

\dagger

uses Reversed New Feedback as the comparison at

N=16

. Brackets give paired 95% user-bootstrap confidence intervals.

	Correct User Profile CQS $\uparrow$				Wrong-User PCVP Gain $\uparrow$
Variant	$\Delta E_{00}$	LPIPS	PSNR	SSIM	$\Delta E_{00}$	LPIPS	PSNR	SSIM
$-$ Explicit Feature $-$ Difference	.20928	21.00842	31.73068	.91467	+.00102	+.10579	+.10323	+.00028
$-$ Set Transformer	.19050	19.72441	29.69401	.91261	-.00014	-.00014	-.00290	-.00001
$-$ Query Feature	.20657	19.77052	33.42625	.91171	+.00081	+.08989	+.04642	+.00028
PrefLUT	.21034	21.20822	31.94439	.91523	+.00410	+.42875	+.38935	+.00139

Appendix

Table F2: Core ablations on PPSD. (A) CQS and Fixed Wrong-User Profile PCVP gains. (B) Differences in Fixed Wrong-User Profile PCVP gain, computed as each variant’s gain minus its matched full model’s gain, with paired 95% user-bootstrap confidence intervals. Blue values mark intervals entirely below zero. Minus signs denote removed components, with the Set Transformer replaced by masked mean pooling.

	Correct User Profile CQS $\uparrow$				Wrong-User PCVP Gain $\uparrow$
Profile Dimension	$\Delta E_{00}$	LPIPS	PSNR	SSIM	$\Delta E_{00}$	LPIPS	PSNR	SSIM
32	.20557	20.8993	31.2043	.91469	.00138	.1432	.1089	.00046
64	.20773	21.0845	31.4524	.91501	.00237	.2263	.2219	.00072
128	.20806	21.0872	31.5174	.91502	.00312	.3105	.2948	.00099
256	.21034	21.2082	31.9444	.91523	.00410	.4287	.3894	.00139
512	.21014	21.1969	31.9005	.91520	.00411	.4284	.3896	.00138

Appendix

Table F3: Reusable User Profile dimension on PPSD. CQS uses the correct user profile, and PCVP gain compares it with the Fixed Wrong-User Profile. The standard profile has 256 dimensions.

Figure H1: Personalized editing comparisons on PPSD. Queries are grouped by user. Input and Preferred show the non-preferred query and its paired preferred target. All methods share queries and 16 reference pairs per user. Labels report direct LPIPS $\downarrow$ /PSNR $\uparrow$ /SSIM $\uparrow$ against Preferred.

Figure H2: PrefLUT PCVP examples on PPSD. Each row compares Correct conditioning with the five controls in Appendix D.1 . Wrong query denotes the Cyclic Wrong-Query Control, which uses another query from the same user to predict the LUT and edit strength applied to the displayed input. Labels report direct LPIPS $\downarrow$ /PSNR $\uparrow$ /SSIM $\uparrow$ against Preferred.

Figure H3: PrefLUT profile refinement on PPSD. Each row fixes the query and network weights while the profile is rebuilt from the first

N=4,8,12,16,24

feedback pairs. Input is the non-preferred query, and Preferred is its paired target. Labels report per-output LPIPS-CQS , PSNR-CQS , and SSIM-CQS (all

\uparrow

). Per-output CQS compares each output with Preferred and Input (Appendix B.4 ).

Explore similar work

HP-Edit: A Human-Preference Post-Training Framework for Image Editing
Apr 21, 2026Fan Li, Chonghuinan Wang, Lina Lei +9Diffusion-Based Image Editing Image Editing
Huawei Noah’s Ark Lab · Nankai University · Harbin Institute of Technology
Learning Color Grading, No Photo Sharing: Federated Aesthetic Preference Learning for Personalized Image Enhancement
Jul 30, 2026Chuanzhi Xu, Ziyuan Tao, Jean Julien KNell +5Image Enhancement Fedavg
The University of Sydney · Macquarie University
Refinement Is Inherently Editable: Training-Free Prompt-to-Prompt Image Editing with Generative Refinement Network
Sep 17, 2026Yulong Chen, Ziqian Zhang, Haoyu Zhang +4Diffusion-Based Image Editing Image Editing
City University of Hong Kong (Dongguan), Guangdong, China · City University of Hong Kong, Hong Kong, China · Mohamed bin Zayed University of Artificial Intelligence, Masdar, Abu Dhabi

Apr 21, 2026, cs.CVJ/K move · Enter open · S save

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

Fan Li, Chonghuinan Wang, Lina Lei, Yuping Qiu+8

Huawei Noah’s Ark Lab · Nankai University · Harbin Institute of Technology

Diffusion-Based Image Editing Image Editing Edit Post-Training

Open paper AskPDF arXiv

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.

Abstract

Explore similar work

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

Learning Color Grading, No Photo Sharing: Federated Aesthetic Preference Learning for Personalized Image Enhancement

Refinement Is Inherently Editable: Training-Free Prompt-to-Prompt Image Editing with Generative Refinement Network