arXiv:2511.18765v1 Announce Type: new 
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.

قدمت NI-Tex نهجًا جديدًا لتوليد قوام الملابس، باستخدام مجموعة بيانات تُدعى 3D Garment Videos لتحسين واقعية القوام المطبق على الشبكات ثلاثية الأبعاد للملابس. تتناول هذه الطريقة القيود المفروضة على التقنيات الحالية التي تتطلب اتساقًا طوبولوجيًا صارمًا بين الصور والشبكات، مما يحسن جودة ومرونة توليد القوام.

NI-Tex ha presentado un enfoque novedoso para la generación de texturas de prendas, utilizando un conjunto de datos llamado 3D Garment Videos para mejorar el realismo de las texturas aplicadas a mallas de prendas 3D. Este método aborda las limitaciones de las técnicas existentes que requieren una consistencia topológica estricta entre las imágenes y las mallas, mejorando así la calidad y flexibilidad de la generación de texturas.

NI-Tex a introduit une nouvelle approche pour la génération de textures de vêtements, utilisant un ensemble de données appelé 3D Garment Videos pour améliorer le réalisme des textures appliquées aux maillages de vêtements 3D. Cette méthode répond aux limitations des techniques existantes qui nécessitent une cohérence topologique stricte entre les images et les maillages, améliorant ainsi la qualité et la flexibilité de la génération de textures.

NI-Tex has introduced a novel approach to garment texture generation, utilizing a dataset called 3D Garment Videos to enhance the realism of textures applied to 3D garment meshes. This method addresses the limitations of existing techniques that require strict topological consistency between images and meshes, thereby improving the quality and flexibility of texture generation.

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

arXiv:2511.19437v1 Announce Type: new 
Abstract: Physically-based rendering (PBR) provides a principled standard for realistic material-lighting interactions in computer graphics. Despite recent advances in generating PBR textures, existing methods fail to address two fundamental challenges: 1) materials decomposition from image prompts under limited illumination cues, and 2) seamless and view-consistent texture completion. To this end, we propose LumiTex, an end-to-end framework that comprises three key components: (1) a multi-branch generation scheme that disentangles albedo and metallic-roughness under shared illumination priors for robust material understanding, (2) a lighting-aware material attention mechanism that injects illumination context into the decoding process for physically grounded generation of albedo, metallic, and roughness maps, and (3) a geometry-guided inpainting module based on a large view synthesis model that enriches texture coverage and ensures seamless, view-consistent UV completion. Extensive experiments demonstrate that LumiTex achieves state-of-the-art performance in texture quality, surpassing both existing open-source and commercial methods.

تم تقديم LumiTex كإطار مبتكر لتوليد القوام القائم على الفيزياء (PBR) عالي الدقة، حيث يعالج التحديات مثل تحليل المواد من إشارات الإضاءة المحدودة وإكمال القوام بشكل سلس. تتضمن هذه الحلول الشاملة مخطط توليد متعدد الفروع، وآلية انتباه للمواد تأخذ في الاعتبار الإضاءة، ووحدة ترميم موجهة بالهندسة لتحسين واقعية القوام.

LumiTex se ha presentado como un marco innovador para la generación de texturas PBR (renderizado basado en la física) de alta fidelidad, abordando desafíos como la descomposición de materiales a partir de indicios de iluminación limitados y la finalización de texturas sin costuras. Esta solución de extremo a extremo incorpora un esquema de generación de múltiples ramas, un mecanismo de atención a materiales consciente de la iluminación y un módulo de inpainting guiado por geometría para mejorar el realismo de las texturas.

LumiTex a été présenté comme un cadre innovant pour la génération de textures PBR (physically-based rendering) de haute fidélité, répondant à des défis tels que la décomposition des matériaux à partir d'indices d'illumination limités et la complétion de textures sans couture. Cette solution de bout en bout intègre un schéma de génération multi-brins, un mécanisme d'attention sur les matériaux conscient de l'éclairage et un module de retouche guidé par la géométrie pour améliorer le réalisme des textures.

LumiTex has been introduced as an innovative framework for high-fidelity physically-based rendering (PBR) texture generation, addressing challenges such as material decomposition from limited illumination cues and seamless texture completion. This end-to-end solution incorporates a multi-branch generation scheme, a lighting-aware material attention mechanism, and a geometry-guided inpainting module to enhance texture realism.

LumiTex: Towards High-Fidelity PBR Texture Generation with Illumination Context

arXiv:2511.16957v1 Announce Type: new 
Abstract: Physically-based rendering (PBR) materials are fundamental to photorealistic graphics, yet their creation remains labor-intensive and requires specialized expertise. While generative models have advanced material synthesis, existing methods lack a unified representation bridging natural image appearance and PBR properties, leading to fragmented task-specific pipelines and inability to leverage large-scale RGB image data. We present MatPedia, a foundation model built upon a novel joint RGB-PBR representation that compactly encodes materials into two interdependent latents: one for RGB appearance and one for the four PBR maps encoding complementary physical properties. By formulating them as a 5-frame sequence and employing video diffusion architectures, MatPedia naturally captures their correlations while transferring visual priors from RGB generation models. This joint representation enables a unified framework handling multiple material tasks--text-to-material generation, image-to-material generation, and intrinsic decomposition--within a single architecture. Trained on MatHybrid-410K, a mixed corpus combining PBR datasets with large-scale RGB images, MatPedia achieves native $1024\times1024$ synthesis that substantially surpasses existing approaches in both quality and diversity.

تم تقديم MatPedia كأساس توليدي عالمي لتوليف المواد عالية الدقة، حيث يعالج التحديات المتعلقة بإنشاء مواد العرض المعتمدة على الفيزياء (PBR) الضرورية للرسوميات الفوتورالية. يستخدم النموذج تمثيلًا مشتركًا جديدًا RGB-PBR يقوم بتشفير المواد في حالتين مترابطتين، مما يسهل إطارًا موحدًا لتوليف المواد.

MatPedia se ha presentado como una base generativa universal para la síntesis de materiales de alta fidelidad, abordando los desafíos de la creación de materiales de renderizado basado en la física (PBR), que son esenciales para gráficos fotorrealistas. El modelo utiliza una nueva representación conjunta RGB-PBR que codifica los materiales en dos latentes interdependientes, facilitando un marco unificado para la síntesis de materiales.

MatPedia a été présenté comme une fondation générative universelle pour la synthèse de matériaux haute fidélité, répondant aux défis de la création de matériaux de rendu basé sur la physique (PBR) essentiels pour des graphismes photoréalistes. Le modèle utilise une nouvelle représentation conjointe RGB-PBR qui encode les matériaux en deux latents interdépendants, facilitant un cadre unifié pour la synthèse de matériaux.

MatPedia has been introduced as a universal generative foundation for high-fidelity material synthesis, addressing the challenges of creating physically-based rendering (PBR) materials which are essential for photorealistic graphics. The model utilizes a novel joint RGB-PBR representation that encodes materials into two interdependent latents, facilitating a unified framework for material synthesis.

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

Was this article worth reading? Share it

TRYONA

Uwear

Aitryon-art