AI 中文总结
本文提出Learning Gaussian Structure（LGS）框架，通过干预引导的高斯结构调整与跨时间点查询，在Waymo Open Dataset和PandaSet上实现了优于现有方法的驾驶场景重建效果。
AI 中文摘要
前向高斯重建是近期兴起的一种高效驾驶场景重建方法。然而，主流基于LiDAR的方法保留了观测点与高斯基元之间的初始对应关系，将初始化的基元集视为最终表示。与基于优化的3DGS不同，这些方法在训练过程中无法累积梯度以确定应如何对场景表示进行致密化。同时，共享的稀疏骨干仅隐式融合不同时间戳的观测结果，未针对单个基元显式聚合跨时间证据。本文提出Learning Gaussian Structure（LGS），这一框架可同时增强高斯结构和基元属性。我们的关键发现是，由剪枝或添加干预引发的局部梯度响应变化，可揭示相应的结构调整是否有利于重建。基于该发现，我们的高斯致密策略从受控干预中学习包含剪枝分数和添加分数的致密图，并在推理过程中直接调整高斯结构。我们进一步开发了紧凑的跨时间点查询，用于显式检索和聚合其他时间戳处高斯基元的邻域特征，以实现可靠的属性预测。在Waymo Open Dataset和PandaSet上开展的大量实验表明，LGS始终优于现有方法。
英文摘要
Feed-forward Gaussian reconstruction has recently emerged as an efficient approach for driving scene reconstruction. However, prevailing LiDAR-based methods preserve the initial correspondence between observed points and Gaussian primitives, treating the initialized primitive set as the final representation. Unlike optimization-based 3DGS, these methods cannot accumulate gradients during training to determine how the scenes representation should be densified. Meanwhile, the shared sparse backbone only fuses observations from different timestamps implicitly, without explicitly aggregating cross-time evidence for individual primitives. In this paper, we present Learning Gaussian Structure (LGS), a framework that enhances both Gaussian structure and primitive attributes. Our key observation is that changes in local gradient responses induced by a prune or add intervention reveal whether the corresponding structural adjustment benefits reconstruction. Based on this observation, our Gaussian Densify Policy learns a Densify Map comprising Prune and Addition Scores from controlled interventions, and directly adjusts the Gaussian structure during inference. We further develop a compact Cross-Time Point Query that explicitly retrieves and aggregates neighboring features from Gaussian primitives at other timestamps for reliable attribute prediction. Extensive experiments on the Waymo Open Dataset and PandaSet demonstrate that LGS consistently outperforms existing methods.
