Radatron: Accurate Detection Using Multi-resolution Cascaded MIMO Radar
Radatron: Accurate Detection Using Multi-resolution Cascaded MIMO Radar
复制标题
DOI：
10.1007/978-3-031-19842-7_10
复制
发表时间：
2022
期刊：
影响因子：
--
通讯作者：
Sohrab Madani;Jayden Guan;Waleed Ahmed;Saurabh Gupta;Haitham Hassanieh
中科院分区：
文献类型：
--
作者：
Sohrab Madani;Jayden Guan;Waleed Ahmed;Saurabh Gupta;Haitham Hassanieh
毫米波（mmWave）雷达由于其在恶劣天气下的有利特性而成为自动驾驶汽车中更受欢迎的感测模态。然而，他们目前缺乏足够的空间分辨率的语义场景理解。在本文中，我们提出了Radatron，一个系统能够准确的目标检测使用毫米波雷达作为一个独立的传感器。为了启用Radatron，我们引入了一个通过级联MIMO（多输入多输出）雷达收集的首个高分辨率汽车雷达数据集。我们的雷达达到5厘米的距离分辨率和1.2角分辨率，比其他公开可用的数据集更精细。我们还开发了一种新的混合雷达处理和深度学习方法，以实现高车辆检测精度。我们对Radatron进行了训练和广泛的评估，以显示它在2D边界框检测中的AP和AP准确性，以及对现有技术的改进。代码和数据集可在https://jguan.page/Radatron/上获得。
Millimeter wave (mmWave) radars are becoming a more popular sensing modality in self-driving cars due to their favorable characteristics in adverse weather. Yet, they currently lack sufficient spatial resolution for semantic scene understanding. In this paper, we present Radatron, a system capable of accurate object detection using mmWave radar as a stand-alone sensor. To enable Radatron, we introduce a first-of-its-kind, high-resolution automotive radar dataset collected with a cascaded MIMO (Multiple Input Multiple Output) radar. Our radar achieves 5 cm range resolution and 1.2angular resolution,finer than other publicly available datasets. We also develop a novel hybrid radar processing and deep learning approach to achieve high vehicle detection accuracy. We train and extensively evaluate Radatron to show it achievesAPandAPaccuracy in 2D bounding box detection, anandimprovement over prior art respectively. Code and dataset is available on https://jguan.page/Radatron/.