用高斯点阵实现像素与标注的快速对应，提升密度回归效率
2D Gaussians Spatial Transport for Point-supervised Density Regression
- 基于高斯点阵建立像素与标注的映射关系
- 相比传统最优传输，训练时无需迭代计算运输方案
- 适用于人群计数、关键点检测等密度回归任务
本文提出高斯空间传输（Gaussian Spatial Transport, GST），利用高斯点阵将图像坐标空间的概率测度传输至标注图。我们设计了一种基于高斯点阵的方法来估计像素与标注之间的对应关系，并据此推导出贝叶斯概率框架下的运输方案。为将该运输方案融入标准网络优化，我们构建了一个衡量运输后差异的损失函数。在人群计数和关键点检测等典型计算机视觉任务上的大量实验验证了该方法的有效性。相比传统最优传输方法，GST 在训练过程中无需迭代计算运输方案，显著提升了效率。代码已开源：https://github.com/infinite0522/GST。
原文摘要 · Abstract (English)
This paper introduces Gaussian Spatial Transport (GST), a novel framework that leverages Gaussian splatting to facilitate transport from the probability measure in the image coordinate space to the annotation map. We propose a Gaussian splatting-based method to estimate pixel-annotation correspondence, which is then used to compute a transport plan derived from Bayesian probability. To integrate the resulting transport plan into standard network optimization in typical computer vision tasks, we derive a loss function that measures discrepancy after transport. Extensive experiments on representative computer vision tasks, including crowd counting and landmark detection, validate the effectiveness of our approach. Compared to conventional optimal transport schemes, GST eliminates iterative transport plan computation during training, significantly improving efficiency. Code is available at https://github.com/infinite0522/GST.
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。