Detecting accurate 3D bounding boxes of the object from point clouds is a major task in autonomous driving perception. At present, the anchor-based or anchor-free models that use LiDAR point clouds for 3D object detection use the center assigner strategy to infer the 3D bounding boxes. However, in the real-world scene, due to the occlusions and the effective detection range of the LiDAR system, only part of the object surface can be covered by the collected point clouds, and there are no measured 3D points corresponding to the physical object center. Obtaining the object by aggregating the incomplete surface point clouds will bring a loss of accuracy in direction and dimension estimation. To address this problem, we propose a corner-guided anchor-free single-stage 3D object detection model (CG-SSD). Firstly, the point clouds within a single frame are assigned to regular 3D grids. 3D sparse convolution backbone network composed of residual layers and sub-manifold sparse convolutional layers are used to construct bird’s eye view (BEV) features for further deeper feature mining by a lite U-shaped network; Secondly, a novel corner-guided auxiliary module (CGAM) with adaptive corner classification algorithm is proposed to incorporate corner supervision signals into the neural network. CGAM is explicitly designed and trained to estimate locations of partially visible and invisible corners to obtain a more accurate object feature representation, especially for small or partial occluded objects; Finally, the deep features from both the backbone networks and CGAM module are concatenated and fed into the head module to predict the classification and 3D bounding boxes of the objects in the scene. The experiments demonstrate that CG-SSD achieves the state-of-art performance on the ONCE benchmark for supervised 3D object detection using single frame point cloud data, with 62.77% mAP. Additionally, the experiments on ONCE and Waymo Open Dataset show that CGAM can be extended to most anchor-based models which use the BEV feature to detect objects, as a plug-in and bring +1.17%∼+14.23% AP improvement. The code is available at https://github.com/mrqrs/CG-SSD.

从点云中检测物体的准确 3D 边界框是自动驾驶感知的一项主要任务。目前，使用 LiDAR 点云进行 3D 对象检测的基于锚点或无锚点的模型使用中心分配器策略来推断 3D 边界框。但在现实场景中，由于LiDAR系统的遮挡和有效探测范围，采集到的点云只能覆盖物体表面的一部分，没有对应物理物体的3D实测点中央。通过聚合不完整的表面点云来获取对象会带来方向和维度估计的准确性损失。为了解决这个问题，我们提出了一种角引导的无锚单级 3D 对象检测模型（CG-SSD）。首先，单个帧内的点云被分配给常规的 3D 网格。由残差层和子流形稀疏卷积层组成的 3D 稀疏卷积骨干网络用于构建鸟瞰图 (BEV) 特征，以便通过精简的 U 形网络进行更深层次的特征挖掘；其次，提出了一种具有自适应角点分类算法的新型角点引导辅助模块（CGAM），将角点监督信号整合到神经网络中。CGAM 被明确地设计和训练来估计部分可见和不可见角的位置，以获得更准确的对象特征表示，特别是对于小的或部分被遮挡的对象；最后，来自骨干网络和 CGAM 模块的深度特征被连接起来并输入到 head 模块中，以预测场景中对象的分类和 3D 边界框。实验表明，CG-SSD 在使用单帧点云数据进行监督 3D 对象检测的 ONCE 基准上实现了最先进的性能，mAP 为 62.77%。此外，在 ONCE 和 Waymo Open Dataset 上的实验表明，CGAM 作为插件可以扩展到大多数使用 BEV 特征检测对象的基于锚的模型，并带来 +1.17%∼+14.23% 的 AP 改进。该代码可在 https://github.com/mrqrs/CG-SSD 获得。实验表明，CG-SSD 在使用单帧点云数据进行监督 3D 对象检测的 ONCE 基准上实现了最先进的性能，mAP 为 62.77%。此外，在 ONCE 和 Waymo Open Dataset 上的实验表明，CGAM 作为插件可以扩展到大多数使用 BEV 特征检测对象的基于锚的模型，并带来 +1.17%∼+14.23% 的 AP 改进。该代码可在 https://github.com/mrqrs/CG-SSD 获得。实验表明，CG-SSD 在使用单帧点云数据进行监督 3D 对象检测的 ONCE 基准上实现了最先进的性能，mAP 为 62.77%。此外，在 ONCE 和 Waymo Open Dataset 上的实验表明，CGAM 作为插件可以扩展到大多数使用 BEV 特征检测对象的基于锚的模型，并带来 +1.17%∼+14.23% 的 AP 改进。该代码可在 https://github.com/mrqrs/CG-SSD 获得。