Train in Germany, Test in the USA: Making 3D Object Detectors Generalize
Train in Germany, Test in the USA: Making 3D Object Detectors Generalize
复制标题
DOI：
10.1109/cvpr42600.2020.01173
复制
发表时间：
2020-05
期刊：
影响因子：
--
通讯作者：
Yan Wang;Xiangyu Chen;Yurong You;Li Erran;B. Hariharan;M. Campbell;Kilian Q. Weinberger;Wei-Lun Chao
中科院分区：
文献类型：
--
作者：
Yan Wang;Xiangyu Chen;Yurong You;Li Erran;B. Hariharan;M. Campbell;Kilian Q. Weinberger;Wei-Lun Chao
在自动驾驶领域，深度学习大大提高了LiDAR和立体相机数据的3D物体检测精度。虽然深度网络非常擅长泛化，但它们也因过度拟合各种虚假伪像而臭名昭著，例如亮度，汽车尺寸和型号，这些伪像可能在整个数据中一致出现。事实上，大多数自动驾驶数据集都是在一个国家的一小部分城市内收集的，通常是在类似的天气条件下。在本文中，我们考虑的任务，适应3D对象检测器从一个数据集到另一个。我们天真地观察到，这似乎是一项非常具有挑战性的任务，导致准确性水平急剧下降。我们提供了大量的实验来调查真正的适应挑战，并得出了一个令人惊讶的结论：要克服的主要适应障碍是不同地理区域的汽车尺寸差异。基于平均汽车尺寸的简单校正产生适应差距的强校正。我们提出的方法很简单，很容易融入大多数3D对象检测框架。它为各国的3D物体检测适应提供了第一个基线，并希望潜在的问题可能比人们希望相信的更容易掌握。我们的代码可以在https：//github上找到。com/cxy1997/3D_adapt_auto_driving.
In the domain of autonomous driving, deep learning has substantially improved the 3D object detection accuracy for LiDAR and stereo camera data alike. While deep networks are great at generalization, they are also notorious to overfit to all kinds of spurious artifacts, such as brightness, car sizes and models, that may appear consistently throughout the data. In fact, most datasets for autonomous driving are collected within a narrow subset of cities within one country, typically under similar weather conditions. In this paper we consider the task of adapting 3D object detectors from one dataset to another. We observe that naively, this appears to be a very challenging task, resulting in drastic drops in accuracy levels. We provide extensive experiments to investigate the true adaptation challenges and arrive at a surprising conclusion: the primary adaptation hurdle to overcome are differences in car sizes across geographic areas. A simple correction based on the average car size yields a strong correction of the adaptation gap. Our proposed method is simple and easily incorporated into most 3D object detection frameworks. It provides a first baseline for 3D object detection adaptation across countries, and gives hope that the underlying problem may be more within grasp than one may have hoped to believe. Our code is available at https://github. com/cxy1997/3D_adapt_auto_driving.