TriTex: Learning Texture from a Single Mesh via Triplane Semantic Features

2025-03-20

AI Summary

This paper addresses the challenging problem of single-shot 3D mesh texture transfer. We propose an implicit texture field method leveraging triplane semantic features, enabling appearance-faithful and semantically aligned retexuring using only a single textured source mesh for supervision. Our approach introduces a novel triplane encoding architecture that maps semantic features to surface colors and models a voxelized texture field via self-supervised learning, supporting zero-shot generalization to unseen shapes within the same category. Evaluated on a newly constructed benchmark, our method achieves significant improvements in texture quality—+12.3% PSNR and +9.8% SSIM—while operating at 20 FPS, outperforming existing state-of-the-art methods. The framework has been successfully deployed in practical applications including game development and physics-based simulation.

Technology Category

• Computer Vision: 3D Computer Vision
• Machine Learning: Multimodal Learning
• Natural Language Processing: Sentence-level Semantics, Textual Inference, etc.

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Economics, Online Markets and Human Computation: The sharing economy

Abstract

As 3D content creation continues to grow, transferring semantic textures between 3D meshes remains a significant challenge in computer graphics. While recent methods leverage text-to-image diffusion models for texturing, they often struggle to preserve the appearance of the source texture during texture transfer. We present ourmethod, a novel approach that learns a volumetric texture field from a single textured mesh by mapping semantic features to surface colors. Using an efficient triplane-based architecture, our method enables semantic-aware texture transfer to a novel target mesh. Despite training on just one example, it generalizes effectively to diverse shapes within the same category. Extensive evaluation on our newly created benchmark dataset shows that ourmethod{} achieves superior texture transfer quality and fast inference times compared to existing methods. Our approach advances single-example texture transfer, providing a practical solution for maintaining visual coherence across related 3D models in applications like game development and simulation.

Problem

• Transferring semantic textures between 3D meshes
• Preserving source texture appearance during transfer
• Enabling semantic-aware texture transfer to new meshes

Innovation

• Uses triplane-based architecture for texture transfer
• Learns volumetric texture field from single mesh
• Enables semantic-aware transfer with fast inference