PrefLUT: Reusable and Refinable Personalized Color Editing from Pairwise Preferences

2026-09-28

AI Summary

This study addresses the limitations of existing methods in modeling persistent user preferences and achieving reusable personalized color editing. To this end, we propose a lightweight user profiling framework based on pairwise preferences. By leveraging a query-conditioned LUT predictor and an edit intensity controller, the method encodes user preferences and maps them into the 3D LUT space to generate exportable color lookup tables, enabling personalized editing without per-user optimization. Furthermore, the model supports dynamic refinement through newly added preference pairs. Experimental results demonstrate that each user profile requires only 260 bytes of storage, with an inference speed of 1.365 milliseconds per image, effectively balancing efficient deployment with enhanced personalization.

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.

Problem

• personalized color editing
• user preference modeling
• lookup table
• pairwise preferences
• reusable user profile

Innovation

• Personalized Color Editing
• 3D LUT
• Preference Modeling
• Reusable User Profile
• PCVP