用知识蒸馏压缩3D高斯点云，大幅降低存储且保持画质。
Distilled-3DGS:Distilled 3D Gaussian Splatting
- 用多个教师模型指导轻量学生模型优化，实现高效知识迁移。
- 在多个数据集上，渲染质量接近顶尖方法，存储量减少超50%。
- 适合需要低内存占用的3D重建与实时渲染场景。
3D高斯点云（3DGS）在新视角合成任务中表现优异，但高质量渲染通常需大量3D高斯分布，导致内存和存储开销巨大。为此，我们提出首个针对3DGS的知识蒸馏框架，包含原始3DGS、噪声增强变体及丢弃正则化版本等多种教师模型。通过聚合教师输出，指导轻量学生模型优化。为保留隐含几何结构，设计结构相似性损失，提升学生与教师在空间分布上的一致性。在多个数据集上的定量与定性评估表明，该方法无需复杂组件，即可在渲染质量与存储效率方面超越现有最优方法。项目主页：https://distilled3dgs.github.io；代码：https://github.com/lt-xiang/Distilled-3DGS。
原文摘要 · Abstract (English)
3D Gaussian Splatting (3DGS) has exhibited remarkable efficacy in novel view synthesis (NVS). However, it suffers from a significant drawback: achieving high-fidelity rendering typically necessitates a large number of 3D Gaussians, resulting in substantial memory consumption and storage requirements. To address this challenge, we propose the first knowledge distillation framework for 3DGS, featuring various teacher models, including vanilla 3DGS, noise-augmented variants, and dropout-regularized versions. The outputs of these teachers are aggregated to guide the optimization of a lightweight student model. To distill the hidden geometric structure, we propose a structural similarity loss to boost the consistency of spatial geometric distributions between the student and teacher model. Through comprehensive quantitative and qualitative evaluations across diverse datasets, the proposed Distilled-3DGS, a simple yet effective framework without bells and whistles, achieves promising rendering results in both rendering quality and storage efficiency compared to state-of-the-art methods. Project page: https://distilled3dgs.github.io . Code: https://github.com/lt-xiang/Distilled-3DGS .
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。