NI-Tex: Non-isometric Image-based Garment Texture Generation

2025-11-23

AI Summary

Existing industrial 3D garment mesh textures suffer from insufficient diversity, while mainstream image-driven methods heavily rely on topology consistency and precise pose alignment, limiting both fidelity and flexibility. Method: We propose a novel paradigm for garment texture synthesis from non-isometric images: (i) constructing the first garment-specific 3D dynamic dataset enabling cross-pose texture learning; (ii) designing an uncertainty-guided iterative baking mechanism for robust multi-view texture fusion; and (iii) introducing Nano Banana—previously unexplored in garment editing—to bypass topological constraints. Our dual-branch feedforward network integrates physics-based simulation data, augmented by view selection and reweighting strategies to optimize texture-space alignment. Results: Experiments demonstrate high-fidelity, physically based rendering (PBR) texture generation under complex deformations and multi-pose configurations, significantly improving texture diversity and synthesis flexibility—achieving industrial-grade design standards.

Technology Category

• Computer Vision: Computational Photography, Image & Video Synthesis
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Web Mining and Content Analysis: Web data generation and simulation
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Existing industrial 3D garment meshes already cover most real-world clothing geometries, yet their texture diversity remains limited. To acquire more realistic textures, generative methods are often used to extract Physically-based Rendering (PBR) textures and materials from large collections of wild images and project them back onto garment meshes. However, most image-conditioned texture generation approaches require strict topological consistency between the input image and the input 3D mesh, or rely on accurate mesh deformation to match to the image poses, which significantly constrains the texture generation quality and flexibility. To address the challenging problem of non-isometric image-based garment texture generation, we construct 3D Garment Videos, a physically simulated, garment-centric dataset that provides consistent geometry and material supervision across diverse deformations, enabling robust cross-pose texture learning. We further employ Nano Banana for high-quality non-isometric image editing, achieving reliable cross-topology texture generation between non-isometric image-geometry pairs. Finally, we propose an iterative baking method via uncertainty-guided view selection and reweighting that fuses multi-view predictions into seamless, production-ready PBR textures. Through extensive experiments, we demonstrate that our feedforward dual-branch architecture generates versatile and spatially aligned PBR materials suitable for industry-level 3D garment design.

Problem

• Generating garment textures from non-isometric images
• Overcoming topological inconsistency between images and 3D meshes
• Creating production-ready PBR textures for diverse garment geometries

Innovation

• Constructed 3D Garment Videos dataset for cross-pose texture learning
• Employed Nano Banana for non-isometric image editing
• Proposed iterative baking method via uncertainty-guided view selection