FabricDiffusion: High-Fidelity Texture Transfer for 3D Garments Generation from In-The-Wild Clothing Images

TL;DR

FabricDiffusion is a novel method that accurately transfers fabric textures from a single clothing image to 3D garments, overcoming occlusion and distortion issues by extracting tileable textures for realistic rendering.

Contribution

We propose a diffusion-based approach that extracts distortion-free, tileable textures from single images, enabling high-fidelity 3D garment texture transfer and realistic relighting.

Findings

Outperforms state-of-the-art methods on synthetic and real-world images.

Effectively captures textures, patterns, and details like logos.

Generalizes well to unseen textures and shapes.

Abstract

We introduce FabricDiffusion, a method for transferring fabric textures from a single clothing image to 3D garments of arbitrary shapes. Existing approaches typically synthesize textures on the garment surface through 2D-to-3D texture mapping or depth-aware inpainting via generative models. Unfortunately, these methods often struggle to capture and preserve texture details, particularly due to challenging occlusions, distortions, or poses in the input image. Inspired by the observation that in the fashion industry, most garments are constructed by stitching sewing patterns with flat, repeatable textures, we cast the task of clothing texture transfer as extracting distortion-free, tileable texture materials that are subsequently mapped onto the UV space of the garment. Building upon this insight, we train a denoising diffusion model with a large-scale synthetic dataset to rectify distortions in the input texture image. This process yields a flat texture map that enables a tight coupling with existing Physically-Based Rendering (PBR) material generation pipelines, allowing for realistic relighting of the garment under various lighting conditions. We show that FabricDiffusion can transfer various features from a single clothing image including texture patterns, material properties, and detailed prints and logos. Extensive experiments demonstrate that our model significantly outperforms state-to-the-art methods on both synthetic data and real-world, in-the-wild clothing images while generalizing to unseen textures and garment shapes.