DiffSoup：面向极端辐射场简化的三角形汤直接可微光栅化
DiffSoup: Direct Differentiable Rasterization of Triangle Soup for Extreme Radiance Field Simplification
用少量带神经纹理与二值不透明度的三角形汤替代数十万高斯球，并通过随机不透明度掩码让二值不透明度直接可微，从而绕开平滑光栅化；对需要把辐射场压到移动端、或想复用传统深度测试图形管线的视觉研究者，这是一条可复用的极端简化与可微渲染思路。
原文依据 this binary opacity representation is directly differentiable via stochastic opacity masking, enabling stable training without a mollifier
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论辐射场重建旨在从多视角 RGB 图像中恢复高质量的三维表示。近期诸如 3D 高斯泼溅等进展，使得在性能足够的图形硬件上能够以高视觉保真度进行实时渲染。然而，要在多种平台上实现高效的在线传输与渲染，需要大幅简化模型，将图元数量减少若干数量级。我们提出 DiffSoup，一种辐射场表示，它采用由少量三角形构成的高度非结构化集合（即三角形汤），并为其赋予神经纹理与二值不透明度。我们证明，这种二值不透明度表示可通过随机不透明度掩码直接求导，从而无需平滑化算子（即平滑光栅化）即可稳定训练。DiffSoup 可以使用标准深度测试进行光栅化，从而无缝集成到传统图形管线中，并在消费级笔记本电脑与移动设备上实现交互式渲染。
英文原文摘要 991 字符
Radiance field reconstruction aims to recover high-quality 3D representations from multi-view RGB images. Recent advances, such as 3D Gaussian splatting, enable real-time rendering with high visual fidelity on sufficiently powerful graphics hardware. However, efficient online transmission and rendering across diverse platforms requires drastic model simplification, reducing the number of primitives by several orders of magnitude. We introduce DiffSoup, a radiance field representation that employs a soup (i.e., a highly unstructured set) of a small number of triangles with neural textures and binary opacity. We show that this binary opacity representation is directly differentiable via stochastic opacity masking, enabling stable training without a mollifier (i.e., smooth rasterization). DiffSoup can be rasterized using standard depth testing, enabling seamless integration into traditional graphics pipelines and interactive rendering on consumer-grade laptops and mobile devices.