Geometry-Aware Texture Generation for 3D Head Modeling with Artist-driven Control (CVPRW 2025)

Amin Fadaeinejad*1,2, Abdallah Dib*1, Luiz Gustavo Hafemann1, Emeline Got1, Trevor Anderson1, Amaury Depierre1, Nikolaus F. Troje2 Marcus A Brubaker2 Marc-AndrΓ© Carbonneau1
1Ubisoft LaForge, 2York University
*Equal contribution
Paper arXiv

We propose a generative head model that gives artists control over facial structure, skin tone variations, and fine-grained details.

Abstract

Creating realistic 3D head assets for virtual characters that match a precise artistic vision remains labor-intensive. We present a novel framework that streamlines this process by providing artists with intuitive control over generated 3D heads. Our approach uses a geometry-aware texture synthesis pipeline that learns correlations between head geometry and skin texture maps across different demographics. The framework offers three levels of artistic control: manipulation of overall head geometry, adjustment of skin tone while preserving facial characteristics, and fine-grained editing of details such as wrinkles or facial hair. Our pipeline allows artists to make edits to a single texture map using familiar tools, with our system automatically propagating these changes coherently across the remaining texture maps needed for realistic rendering. Experiments demonstrate that our method produces diverse results with clean geometries. We showcase practical applications focusing on intuitive control for artists, including skin tone adjustments and simplified editing workflows for adding age-related details or removing unwanted features from scanned models. This integrated approach aims to streamline the artistic workflow in virtual character creation.

Pipeline

Dataset samples

Overview of the proposed pipeline: A generator 𝓕 generates a mesh. Another generator G generates an intermediate skin representation: a melanin map π“œ and a high-frequency details map 𝓗. G𝓐 generates a color map 𝓐 from π“œ, allowing for precise control of skin tone. G𝓒 generates a final skin reflectance map 𝓒, incorporating arbitrary manipulations on 𝓗. GΞ΅ decomposes the intrinsic face reflectance maps, which can be used to render realistic heads.

Video

Our Contributions

• A multi-level control framework for 3D head generation that enables artists to manipulate geometry, skin tone, and fine details independently.

• A geometry-aware texture generation approach that maintains consistency between facial structure and appearance.

• A novel approach for editing fine-grained details in intrinsic texture maps. This allows artists to modify a single texture map using any image editing tool, with the changes coherently propagated to the intrinsic texture maps.

• A data-driven method for precise manipulation of skin tones while preserving other facial attributes, enabling greater diversity in virtual character creation.

BibTeX

@article{Fadaeinejad_2025_cveu,
  author    = {Amin Fadaeinejad, Abdallah Dib, Luiz Gustavo Hafemann, Emeline Got, Trevor Anderson, Amaury Depierre, Nikolaus F. Troje, Marcus A Brubaker, Marc-AndrΓ© Carbonneau},
  title     = {Geometry-Aware Texture Generation for 3D Head Modeling with Artist-driven Control},
  journal   = {AI for Creative Visual Content Generation Editing and Understanding (CVEU), CVPR Workshop},
  year      = {2025},
}