English

Seed3D 2.0: Advancing High-Fidelity Simulation-Ready 3D Content Generation

Graphics 2026-05-15 v1 Computer Vision and Pattern Recognition Image and Video Processing

Abstract

We present Seed3D 2.0, an advanced 3D content generation system built on Seed3D 1.0, with substantial improvements across generation fidelity, simulation-ready capabilities, and application coverage. For geometry, a coarse-to-fine two-stage pipeline decouples global structure learning from high-frequency detail recovery, while a locality-aware VAE achieves higher spatial compression and more efficient decoding. For texture and material generation, we replace the cascaded pipeline of Seed3D 1.0 with a unified PBR model that directly generates multi-view albedo and metallic-roughness maps, enhanced by Mixture-of-Experts scaling and VLM-based semantic conditioning for improved material precision and visual fidelity. Beyond single-object generation, Seed3D 2.0 introduces a simulation-ready model suite comprising scene layout planning, part-aware decomposition, and training-free articulation generation, enabling coherent scene construction and part-level physical interaction across physics and graphics engines. A large-scale human preference study against five recent commercial models shows that Seed3D 2.0 achieves consistent win rates of 69.0% to 89.9% in textured 3D asset generation. Seed3D 2.0 is available on https://exp.volcengine.com/ark/vision?_vtm_=0.0.c70961.d701978.0&mode=vision&modelId=doubao-seed3d-2-0-260328&tab=Gen3D

Keywords

Cite

@article{arxiv.2605.13862,
  title  = {Seed3D 2.0: Advancing High-Fidelity Simulation-Ready 3D Content Generation},
  author = {Diandian Gu and Jing Lin and Gaohong Liu and Jiahang Liu and Su Ma and Guang Shi and Jun Wang and Qinlong Wang and Qianyi Wu and Zhongcong Xu and Xuanyu Yi and Zihao Yu and Jianfeng Zhang and Zhuolin Zheng and Yifan Zhu and Rui Chen and Hengkai Guo and Xiaoyang Guo and Mingcong Han and Xu Han and Xiu Li and Yixun Liang and Weiqiang Lou and Junzhe Lu and Guan Luo and Minghan Qin and Shuguang Wang and Yuang Wang},
  journal= {arXiv preprint arXiv:2605.13862},
  year   = {2026}
}

Comments

Seed3D 2.0 Technical Report; Official Page on https://seed.bytedance.com/seed3d_2_0