Abstract
Image-to-3D generation has become increasingly capable of producing objects that closely resemble the input image, and an outstanding challenge is to reproduce the depicted object itself, including the specific geometry that defines it. Inscriptions, brand marks, and repeated structures are frequently distorted or lost, despite being critical to object identity. We present Hi3D 3.0, an image-to-3D generation system targeting object-specific fidelity, with Twinkle3D as its geometry model for generating watertight triangle meshes at $2048^{3}$ resolution. Twinkle3D advances high-fidelity geometry generation along four dimensions. First, while O-Voxel/FaithC offers high representational precision, it often suffers from poor surface quality and non-watertight geometry. We address both issues while retaining its $2048^{3}$-level precision. Second, we scale diffusion generation to sequences of up to 300K geometric tokens through a redesigned DiT architecture and large-scale distributed training optimizations, reducing training time per step from approximately ten minutes to ten seconds. Third, subsequent refinement cannot fully compensate for errors introduced during initial generation; we therefore strengthen both global shape and local detail in the initial generation stage, and the resulting single-stage model surpasses prior two-stage pipelines with $512^{3}$ refinement. Finally, we introduce a fine-grained image-3D cross-modal interaction mechanism that strengthens correspondence between visual evidence and geometric tokens, improving the recovery of object-specific structures. We evaluate geometric fidelity using alignment metrics derived from silhouettes and normal fields. Hi3D 3.0 outperforms four commercial systems across all reported metrics, recovering 82.1% of inscribed characters at 98.2% precision, compared with 21.7% recall for the strongest competitor.
Keywords
Subject
Publication details
- Journal
- Not available
- Open access
- Green open access
Cite this article
APA 7
Li, Z., Zhao, S., He, H., Chen, Y., Huang, J., Li, B., Li, C., Li, J., Li, J., Liu, R., Luo, Y., Sun, T., Tang, P., Wang, S., Wu, J., Wu, K., Yang, K., Yang, Z., Zhang, H., Zhao, X., Zheng, X., Li, Y., Zheng, H., & Jia, R. (2026). HI3D 3.0 (Twinkle3D): Object-specific 3D Asset Generation with High Resolution. https://omanscience.com/en/articles/hi3d-3-0-twinkle3d-object-specific-3d-asset-generation-with-high-resolution
MLA 9
Li, Ziying, et al. "HI3D 3.0 (Twinkle3D): Object-specific 3D Asset Generation with High Resolution." https://omanscience.com/en/articles/hi3d-3-0-twinkle3d-object-specific-3d-asset-generation-with-high-resolution.
Chicago (author–date)
Li, Ziying, Shengchu Zhao, Huiang He, Yiyang Chen, Jianwen Huang, Bailin Li, Changhao Li, Jianhui Li, Jie Li, Ruiyang Liu, Yibo Luo, Tengjiao Sun, Pei Tang, Shiwen Wang, Jiaqi Wu, Kang Wu, Kaiqiao Yang, Zherui Yang, Hu Zhang, Xuezhi Zhao, Xinhe Zheng, Yukun Li, Heliang Zheng, and Rongfei Jia. 2026. "HI3D 3.0 (Twinkle3D): Object-specific 3D Asset Generation with High Resolution." https://omanscience.com/en/articles/hi3d-3-0-twinkle3d-object-specific-3d-asset-generation-with-high-resolution.
Harvard
Li, Z., Zhao, S., He, H., Chen, Y., Huang, J., Li, B., Li, C., Li, J., Li, J., Liu, R., Luo, Y., Sun, T., Tang, P., Wang, S., Wu, J., Wu, K., Yang, K., Yang, Z., Zhang, H., Zhao, X., Zheng, X., Li, Y., Zheng, H. and Jia, R. (2026) 'HI3D 3.0 (Twinkle3D): Object-specific 3D Asset Generation with High Resolution', Available at: https://omanscience.com/en/articles/hi3d-3-0-twinkle3d-object-specific-3d-asset-generation-with-high-resolution.
Vancouver
Li Z, Zhao S, He H, Chen Y, Huang J, Li B, et al. HI3D 3.0 (Twinkle3D): Object-specific 3D Asset Generation with High Resolution. https://omanscience.com/en/articles/hi3d-3-0-twinkle3d-object-specific-3d-asset-generation-with-high-resolution
IEEE
Z. Li, S. Zhao, H. He, Y. Chen, J. Huang, B. Li, C. Li, J. Li, J. Li, R. Liu, Y. Luo, T. Sun, P. Tang, S. Wang, J. Wu, K. Wu, K. Yang, Z. Yang, H. Zhang, X. Zhao, X. Zheng, Y. Li, H. Zheng, and R. Jia, "HI3D 3.0 (Twinkle3D): Object-specific 3D Asset Generation with High Resolution," https://omanscience.com/en/articles/hi3d-3-0-twinkle3d-object-specific-3d-asset-generation-with-high-resolution.