Seed3D 1.0: From Images to High-Fidelity Simulation-Ready 3D Assets Model ID: doubao-seed3d-1-0-250928
2025
Read Seed3D 1.0: From Images to High-Fidelity Simulation-Ready 3D Assets Model ID: doubao-seed3d-1-0-250928 (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-2510-19944.txt.
Description
Jiashi Feng and twenty-seven co-authors present Seed3D 1.0, a foundation model that turns a single image into a three-dimensional asset ready to drop into a physics simulator. Embodied AI needs training worlds that are both varied and physically accurate; video generators give variety without interactive physics, while physics engines need costly hand-built assets. Seed3D targets the gap.
Its assets carry accurate geometry, well-aligned textures, and realistic physically based materials, and they enter physics engines with minimal setup for robotics manipulation and simulation training. The system also assembles individual objects into coherent full scenes. The authors frame the release as infrastructure for scalable, physics-based world simulators.
Catalog record
| YEAR | 2025 |
|---|---|
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 61 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2510-19944