Multimodal 3D LUT Generation via StatLUT with Statistical Features for Photorealistic Style Transfer

Jul 16, 2026

Photorealistic Style Transfer transfers color and tonal style from a reference to a content image while preserving spatial structure. Existing deep-learning approaches entangle semantics from pre-trained encoders, causing spatial distortions; pixel-level LUT mapping also overlooks color gamut topology, leading to color banding, and lacks multimodal text control. StatLUT introduces a Lab-Extractor for spatiallyagnostic statistical features, a Transformer Seq2Seq LUT generator with an MR-Mapper for topology-smooth LUTs, and an H-Diffuser diffusion transformer to synthesize features from natural language prompts, achieving robust multimodal quality gains.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages17

Reading Time43 min

Introduction

Contributions

Related Work

Feature-Encoding-based Photorealistic Style Transfer

Pixel-Level Mapping and 3D LUT Generation

arXiv :2607 .08227v 1 [ cs .CV] 9 Jul 2026 Multimodal 3D LUT Generation via StatLUT with Statistical Features for Photorealistic Style Transfer Yifan Wang, Zhixiang Hao, Yu Wang Congchao Zhu Honor Device Co., Ltd. [wyf141060@163.com](wyf141060@163.com) , {haozhixiang, wangyu24, [zhucongchao}@honor.com](zhucongchao}@honor.com) Abstract Photorealistic Style Transfer (PST) aims to transfer the color and tonal style of a reference to a content image while strictly preserving its structural integrity. However, existing deep learning-based methods inherently suffer from semantic entanglement caused by pre-trained image encoders, leading to unnatural spatial distortions. Moreover, current pixel-level mapping paradigms often ignore color gamut topology, resulting in color banding, while also lacking the multimodal capability for intuitive text-driven control. To address these bottlenecks, we propose StatLUT, an innovative multimodal framework for 3D LUT generation. First, we bypass traditional encoders and introduce a Lab-Extractor to derive spatiallyagnostic statistical features, fundamentally decoupling color distributions from structural semantics to ensure artifact-free rendering. Second, we formulate LUT generation as a Transformer-based Seq2Seq translation task, utilizing a Multidimensional Residual Mapper (MR-Mapper) to predict topologically smooth 3D LUTs. Finally, to break the single-modal barrier, we propose the H-Diffuser, a lightweight Diffusion Transformer that directly synthesizes statistical features from natural language prompts, enabling flexible text-driven color grading. Extensive experiments on standard benchmarks demonstrate that StatLUT significantly outperforms state-of-the-art methods in both visual quality and quantitative metrics, pioneering a highly robust and flexible paradigm for multimodal photorealistic style transfer. 1 Introduction Photorealistic Style Transfer (PST) aims to faithfully transfer the color and tonal style of a reference image to a content image while strictly preserving the original spatial structure. Unlike artistic style transfer[1–5], PST demands visual realism and structural fidelity, precluding unnatural artifacts or texture distortions during non-linear color mapping. Although deep learning has advanced PST, adhering to this “artifact-free” baseline remains challenging. Current mainstream methods widely adopt pre-trained encoder-decoder architectures[6–8] . However, these models naturally extract high-level semantic features, creating a fundamental mechanistic mismatch with the PST task, which inherently relies on low-level color distributions. This semantic entanglement allows local structural semantics to interfere with color mapping, leading to spatial distortions. Furthermore, their massive memory consumption constitutes a prohibitive bottleneck for high-resolution (e.g., 4K/8K) deployment[9, 10] . To overcome these barriers, recent research has shifted towards a pixel-level mapping paradigm[9– 11], with 3D lookup tables (LUTs) emerging as the standard due to their non-linear capacity and low computational cost[10] . Nevertheless, existing LUT generation mechanisms exhibit significant ∗ Corresponding author. Preprint. limitations: 1) traditional blending paradigms[12, 13] restrict representations to a linear subspace, hindering out-of-distribution generalization; 2) recent generative paradigms[14] fail to completely decouple semantics from color, risking texture artifacts; 3) point-wise mapping approaches[9] ignore the topology between adjacent color bins, causing color banding and inter-frame flickering in videos. Moreover, they lack multimodal capabilities for intuitive text-driven color control. To address these limitations, we propose StatLUT, a novel multimodal photorealistic style transfer and color generation framework. In summary, our main contributions are as follows: • Bypassing traditional image encoders, we utilize multi-dimensional l

FAQs

What problem does StatLUT address in photorealistic style transfer?

It targets semantic entanglement from pre-trained encoders that causes spatial distortions, as well as limitations of LUT mapping that ignore color gamut topology and lack multimodal text-driven control.

How does StatLUT decouple color distributions from structural semantics?

It bypasses traditional encoders and uses a Lab-Extractor to derive spatiallyagnostic low-level statistical features, serving as explicit color priors to block semantic interference.

How does StatLUT enable text-driven color grading without a reference image?

It introduces an H-Diffuser, a lightweight Diffusion Transformer that directly synthesizes statistical features from natural language prompts, breaking the single-modal barrier.

Jul 16, 2026

Upload by Theodora