面向路侧激光雷达场景补全的概率差异学习
Probabilistic Discrepancy Learning for Roadside LiDAR Scene Completion
用扩散模型在概率位姿校正后做全场景点云补全，并设计区域/全局采样差异损失，可迁移到固定视角遮挡下的点云补全与多模态融合感知。
原文依据 utilizing a diffusion model within scene discrepancy learning for robust full-scene completion
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论本文提出一种面向路侧激光雷达场景补全的概率差异学习方法（PDL）。传统方法主要关注目标级补全以及自车视角下的场景补全，难以应对路侧传感器固定视角所导致的长时或严重遮挡。为解决该问题，我们引入外部视觉信息来补偿被遮挡的路侧点云。具体而言，PDL 主要分为概率位姿差异最小化与场景差异学习两部分。我们采用概率位姿差异最小化来校正基于视觉的检测器输出的含噪位姿，同时在场景差异学习中利用扩散模型实现鲁棒的全场景补全。此外，我们设计了区域与全局采样差异学习损失，以实现鲁棒且高效的训练。我们在 V2X-Seq 与 TUMTraf-V2X 路侧数据集上开展了大量实验。结果表明，PDL 达到了最优性能，与现有方法相比，倒角距离（CD）平均降低 14.5%，3D 詹森-香农散度（JSD）平均降低 6%。
英文原文摘要 1173 字符
We propose a probabilistic discrepancy learning approach for roadside LiDAR scene completion (PDL). Conventional methods focus on object-level completion and scene completion from ego-vehicle viewpoint. These methods struggle to cope with long-term or severe occlusions caused by roadside sensors with fixed viewpoints. To address this issue, we compensate for occlusion roadside point clouds by introducing external visual information. Specifically, the PDL is mainly divided into probabilistic pose discrepancy minimization and scene discrepancy learning. We employ probabilistic pose discrepancy minimization to correct noisy poses from vision-based detectors, while utilizing a diffusion model within scene discrepancy learning for robust full-scene completion. Furthermore, we design regional and global sampling discrepancy learning losses to achieve robust and efficient training. We conducted extensive experiments on the V2X-Seq and TUMTraf-V2X roadside datasets. Results demonstrate that PDL achieves state-of-the-art performance, with average reductions of 14.5% in chamfer distance (CD) and 6% in 3D Jensen Shannon divergence (JSD) compared to existing methods.