Pandora3D: A Comprehensive Framework for High-Quality 3D Shape and Texture Generation

Jiayu Yang, Taizhang Shang, Weixuan Sun, Xibin Song, Ziang Cheng, Senbo Wang, Shenzhou Chen, Weizhe Liu, Hongdong Li, Pan Ji · 2025

Read Pandora3D: A Comprehensive Framework for High-Quality 3D Shape and Texture Generation by Jiayu Yang, Taizhang Shang, Weixuan Sun, Xibin Song, Ziang Cheng, Senbo Wang, Shenzhou Chen, Weizhe Liu, Hongdong Li, Pan Ji (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-14247.txt.

Description

Jiayu Yang and co-authors present a 2025 report on generating three-dimensional content with AI. Designers want full objects with surface detail from a photo, several views, or a line of text. The framework, Pandora3D, splits the job into shape building and texture painting, each with its own pipeline.

Shape comes from a model that encodes solid forms into a compact code and grows new codes by diffusion. A mesh-based route is also tried for simpler forms. Texture work starts with a front view, expands to many views, and converts color into surface material maps. A scheduler keeps the views pixel-consistent as they form. Code and trained weights are released in public for others to build on.

Catalog record

CREATORJiayu Yang, Taizhang Shang, Weixuan Sun, Xibin Song, Ziang Cheng, Senbo Wang, Shenzhou Chen, Weizhe Liu, Hongdong Li, Pan Ji
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE35 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2502-14247

Browse the full library · About the Tower of Babel