**摘要**
Sparse voxel representation has emerged as a scalable foundation for image-to-3D Gaussian Splatting (3DGS) generation, yet current methods struggle to preserve high-frequency visual details of input images due to two structural bottlenecks. First, they adopt discriminative 2D features optimized for semantic abstraction to construct sparse voxel latents, which suppress reconstructive cues and induc
👤 作者: Haorui Ji, Weizhe Liu, Hongdong Li, Hengkai Guo

---
🔗 **[FLUX3D ：具有扩散对准稀疏表示的高保真3D高斯生成](https://arxiv.org/abs/2606.24874v1)**

> FLUX3D: High-Fidelity 3D Gaussian Generation with Diffusion-Aligned Sparse Representation
🏷️ 来源: ArXiv cs.AI
⏱️ 2026-06-24 14:00