Year: 2025
Type: conference-paper
Abstract: Figure 1: Text2Texture generates an image from a text prompt and converts this into a textured 3D object ready for 3D printing.(A) Users enter a text prompt in the web tool or upload an image directly.(B) Users can [2] generate or draw labeled masks such as "tree" which synthesizes a texture applied locally to the mask.[3] Users then select scaling parameters that a depth estimator uses to extrude the 2D image into a 2.5D surface.(C) Text2Texture then superimposes the texture and depth map to render a 3D-printable model in the viewport.(D) The 3D printed model captures both macroscopic depth and locally applied textures.
Authors: Joshua Yin, Faraz Faruqi, Martin Nisser
Language: English
Subfield: Computational Mechanics
Field: Engineering
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W7084059838