提出空间自适应正则化，让3D高斯点云更精准地压缩。
SVR-GS: Spatially Variant Regularization for Probabilistic Masks in 3D Gaussian Splatting
- 按每像素贡献动态生成掩码，只在低重要性点施加稀疏压力
- 相比MaskGS减少1.79倍点数，比原版3DGS少5.63倍，画质损失仅0.5dB
- 适合机器人、AR/VR等实时应用，模型更小更快更省内存
3D高斯点云（3DGS）可实现快速高质量的新视角合成，但通常需通过密集化与剪枝优化高斯点数量。现有基于掩码的剪枝方法（如MaskGS）对掩码全局均值进行正则化，与决定图像质量的局部逐像素（逐射线）重建损失不匹配。本文提出SVR-GS，一种空间可变正则化方法，从每个高斯点沿射线的有效贡献中渲染出逐像素空间掩码，从而在低重要性高斯点上施加稀疏压力。我们探索三种空间掩码聚合策略，使用CUDA实现，并进行梯度分析以支持最终设计。在Tanks&Temples、Deep Blending和Mip-NeRF360数据集上的大量实验表明，平均而言，所提方法相较MaskGS将高斯点数减少1.79倍，相较3DGS减少5.63倍，同时画质下降仅0.50 dB和0.40 dB。这些提升使模型更小、更快、更节省内存，非常适合机器人、AR/VR及移动感知等实时应用场景。
原文摘要 · Abstract (English)
3D Gaussian Splatting (3DGS) enables fast, high-quality novel view synthesis but typically relies on densification followed by pruning to optimize the number of Gaussians. Existing mask-based pruning, such as MaskGS, regularizes the global mean of the mask, which is misaligned with the local per-pixel (per-ray) reconstruction loss that determines image quality along individual camera rays. This paper introduces SVR-GS, a spatially variant regularizer that renders a per-pixel spatial mask from each Gaussian's effective contribution along the ray, thereby applying sparsity pressure where it matters: on low-importance Gaussians. We explore three spatial-mask aggregation strategies, implement them in CUDA, and conduct a gradient analysis to motivate our final design. Extensive experiments on Tanks\&Temples, Deep Blending, and Mip-NeRF360 datasets demonstrate that, on average across the three datasets, the proposed SVR-GS reduces the number of Gaussians by 1.79\(\times\) compared to MaskGS and 5.63\(\times\) compared to 3DGS, while incurring only 0.50 dB and 0.40 dB PSNR drops, respectively. These gains translate into significantly smaller, faster, and more memory-efficient models, making them well-suited for real-time applications such as robotics, AR/VR, and mobile perception.
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。