用虚拟点增强远距离激光点云，提升3D目标检测精度。
Point Virtual Transformer
- 融合真实激光点与精选虚拟点，通过变压器建模上下文关系。
- 在KITTI数据集上实现91.16% 3D AP，远超传统方法。
- 适合需要长距感知的自动驾驶系统研发人员参考。
基于激光雷达的3D目标检测器在远距离因点云稀疏而难以检测物体，导致几何线索不足。现有方法通过RGB图像生成深度补全的虚拟点来增强数据，但直接使用全部虚拟点会增加计算开销，并带来真实与虚拟信息融合困难的问题。本文提出Point Virtual Transformer（PointViT），一种基于变压器的3D目标检测框架，联合推理原始激光点与有选择地采样的虚拟点。研究了从早期点级融合到基于鸟瞰图（BEV）的门控融合等多种融合策略，分析其在准确率与效率间的权衡。融合后的点云经体素化并使用稀疏卷积编码为BEV表示，从中初始化一组高置信度的目标查询，并通过变压器上下文聚合模块进行优化。在KITTI基准测试中，汽车类别达到91.16% 3D AP、95.94% BEV AP，以及99.36% 2D检测AP。
原文摘要 · Abstract (English)
LiDAR-based 3D object detectors often struggle to detect far-field objects due to the sparsity of point clouds at long ranges, which limits the availability of reliable geometric cues. To address this, prior approaches augment LiDAR data with depth-completed virtual points derived from RGB images; however, directly incorporating all virtual points leads to increased computational cost and introduces challenges in effectively fusing real and virtual information. We present Point Virtual Transformer (PointViT), a transformer-based 3D object detection framework that jointly reasons over raw LiDAR points and selectively sampled virtual points. The framework examines multiple fusion strategies, ranging from early point-level fusion to BEV-based gated fusion, and analyses their trade-offs in terms of accuracy and efficiency. The fused point cloud is voxelized and encoded using sparse convolutions to form a BEV representation, from which a compact set of high-confidence object queries is initialised and refined through a transformer-based context aggregation module. Experiments on the KITTI benchmark report 91.16% 3D AP, 95.94% BEV AP, and 99.36% AP on the KITTI 2D detection benchmark for the Car class.
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。