Style3D's GarmageNet Bridges Generative AI and Product Workflows in Fashion

Style3D's new AI framework, GarmageNet, connects 2D patterns with 3D garment forms, enabling a shift from isolated image generation to integrated product development workflows.

SA Metrowire Staff
Technology

Style3D, a company specializing in 3D and AI technologies for fashion, is advancing the industry's use of artificial intelligence by moving beyond content generation to connected product workflows. The company recently highlighted its research in a paper presented at SIGGRAPH Asia 2025, demonstrating how AI can integrate design, development, and production processes.

At the core of this advancement is GarmageNet, a framework developed by Style3D Research in collaboration with Zhejiang University, Shanghai Jiao Tong University, and Zhejiang Sci-Tech University. Published in ACM Transactions on Graphics, GarmageNet addresses a central challenge in fashion AI: connecting 2D patterns with 3D garment forms. The framework captures pattern outlines, 3D geometry, and sewing information in a unified representation, allowing it to accept inputs such as text, sketches, images, patterns, or point clouds. It then generates structured assets, reconstructs sewing relationships, and initializes physical simulation for design exploration, digital sampling, and editing.

By linking inputs with patterns, sewing relationships, and simulation, GarmageNet illustrates how fashion AI can become part of a connected product-development workflow rather than remain an isolated image-generation tool. The framework is supported by GarmageSet, a dataset of 14,801 professionally created garments. In evaluations, GarmageJigsaw achieved 99.16% precision and 97.13% recall in sewing-point identification, while GarmageNet achieved a 91.41% simulation-initialization success rate on 150 complex patterns. The reported inference time is approximately eight seconds per garment under the study's test conditions.

The research reflects Style3D's foundation in deformable-object simulation, garment deformation, digital fabrics, and AI-assisted workflows. The company maintains a substantial, dedicated R&D organization and has published more than 30 papers at conferences including SIGGRAPH, SIGGRAPH Asia, CVPR, NeurIPS, and 3DV. As of 28 July 2026, Style3D held 106 granted patents worldwide, including 58 invention patents and 12 granted in Europe and the United States, with a further 155 invention patent applications pending. It also operates an overseas R&D center in Germany.

Style3D's AI and 3D technologies are already being applied across diverse customer workflows. German menswear brand OLYMP uses Style3D and Assyst to connect 2D patterns, 3D simulation, visualization, and digital showrooms. The company reports developing styles in days rather than weeks, reducing physical samples, and improving communication across design, sales, and production.

As Style3D CEO Eric Liu told Just Style, "The most important development is the agent. We use agents to link different skills." Style3D is advancing a connected fashion workflow in which AI agents can coordinate product data, 3D design, simulation, content creation, and production preparation—while human teams retain creative and operational responsibility.

Blockchain Registration

QR Code for Blockchain Registration