Abstract:Texturing is a crucial step in the 3D asset production workflow, which enhances the visual appeal and diversity of 3D assets. Despite recent advancements in Text-to-Texture (T2T) generation, existing methods often yield subpar results, primarily due to local discontinuities, inconsistencies across multiple views, and their heavy dependence on UV unwrapping outcomes. To tackle these challenges, we propose a novel generation-refinement 3D texturing framework called MVPaint, which can generate high-resolution, seamless textures while emphasizing multi-view consistency. MVPaint mainly consists of three key modules. 1) Synchronized Multi-view Generation (SMG). Given a 3D mesh model, MVPaint first simultaneously generates multi-view images by employing an SMG model, which leads to coarse texturing results with unpainted parts due to missing observations. 2) Spatial-aware 3D Inpainting (S3I). To ensure complete 3D texturing, we introduce the S3I method, specifically designed to effectively texture previously unobserved areas. 3) UV Refinement (UVR). Furthermore, MVPaint employs a UVR module to improve the texture quality in the UV space, which first performs a UV-space Super-Resolution, followed by a Spatial-aware Seam-Smoothing algorithm for revising spatial texturing discontinuities caused by UV unwrapping. Moreover, we establish two T2T evaluation benchmarks: the Objaverse T2T benchmark and the GSO T2T benchmark, based on selected high-quality 3D meshes from the Objaverse dataset and the entire GSO dataset, respectively. Extensive experimental results demonstrate that MVPaint surpasses existing state-of-the-art methods. Notably, MVPaint could generate high-fidelity textures with minimal Janus issues and highly enhanced cross-view consistency.

TexPro: Text-guided PBR Texturing with Procedural Material Modeling

TexGen: Text-Guided 3D Texture Generation with Multi-view Sampling and Resampling

Paint3D: Paint Anything 3D with Lighting-Less Texture Diffusion Models

MaPa: Text-driven Photorealistic Material Painting for 3D Shapes

TEXTure: Text-Guided Texturing of 3D Shapes

A Procedural Texture Generation Framework Based on Semantic Descriptions.

A procedural texture generation method based on genetic programming

Text2Tex: Text-driven Texture Synthesis via Diffusion Models

TexGaussian: Generating High-quality PBR Material via Octree-based 3D Gaussian Splatting

Survey of Procedural Methods for Two-Dimensional Texture Generation

TexPainter: Generative Mesh Texturing with Multi-view Consistency

Visual Perception Of Procedural Textures: Identifying Perceptual Dimensions And Predicting Generation Models

An Inverse Procedural Modeling Pipeline for SVBRDF Maps

Learning Pseudo 3D Guidance for View-consistent Texturing with 2D Diffusion

Perception-driven Procedural Texture Generation from Examples.

FlexiTex: Enhancing Texture Generation with Visual Guidance

Toward a Psychophysical-Based Procedural Texture Generation System for Interactive Design

MVPaint: Synchronized Multi-View Diffusion for Painting Anything 3D

TexRO: Generating Delicate Textures of 3D Models by Recursive Optimization

RoomTex: Texturing Compositional Indoor Scenes via Iterative Inpainting

Boosting 3D Object Generation through PBR Materials