Unsupervised Adaptation from Repeated Traversals for Autonomous Driving
Unsupervised Adaptation from Repeated Traversals for Autonomous Driving
复制标题
DOI：
10.48550/arxiv.2303.15286
复制
发表时间：
2023-03
期刊：
影响因子：
--
通讯作者：
Yurong You;Cheng Perng Phoo;Katie Luo;Travis Zhang;Wei-Lun Chao;Bharath Hariharan;Mark E. Campbell;Kilian Q. Weinberger
中科院分区：
文献类型：
--
作者：
Yurong You;Cheng Perng Phoo;Katie Luo;Travis Zhang;Wei-Lun Chao;Bharath Hariharan;Mark E. Campbell;Kilian Q. Weinberger
为了使自动驾驶汽车可靠地运行，其感知系统必须推广到最终用户的环境-理想情况下没有额外的注释工作。一种潜在的解决方案是利用未标记的数据（例如，未标记的LiDAR点云），以使系统适应训练和测试环境之间的差异。虽然对这种无监督域自适应问题进行了广泛的研究，但一个基本问题仍然存在：目标域中没有可靠的信号来监督自适应过程。为了克服这个问题，我们观察到，很容易从重复路线的多次遍历中收集无监督数据。虽然与传统的无监督域自适应不同，但这种假设非常现实，因为许多驾驶员共享相同的道路。我们表明，这个简单的附加假设足以获得一个有效的信号，使我们能够在目标域上执行3D对象检测器的迭代自训练。具体地说，我们生成伪标签与域外检测器，但减少误报，通过删除检测到的假定移动的对象是持久的遍历。此外，我们通过鼓励在不持久的区域进行预测来减少假阴性。我们在两个大规模的驾驶数据集上对我们的方法进行了实验，并在汽车，行人和骑自行车的人的3D对象检测方面取得了显着的改进，使我们更接近可推广的自动驾驶。
For a self-driving car to operate reliably, its perceptual system must generalize to the end-user's environment -- ideally without additional annotation efforts. One potential solution is to leverage unlabeled data (e.g., unlabeled LiDAR point clouds) collected from the end-users' environments (i.e. target domain) to adapt the system to the difference between training and testing environments. While extensive research has been done on such an unsupervised domain adaptation problem, one fundamental problem lingers: there is no reliable signal in the target domain to supervise the adaptation process. To overcome this issue we observe that it is easy to collect unsupervised data from multiple traversals of repeated routes. While different from conventional unsupervised domain adaptation, this assumption is extremely realistic since many drivers share the same roads. We show that this simple additional assumption is sufficient to obtain a potent signal that allows us to perform iterative self-training of 3D object detectors on the target domain. Concretely, we generate pseudo-labels with the out-of-domain detector but reduce false positives by removing detections of supposedly mobile objects that are persistent across traversals. Further, we reduce false negatives by encouraging predictions in regions that are not persistent. We experiment with our approach on two large-scale driving datasets and show remarkable improvement in 3D object detection of cars, pedestrians, and cyclists, bringing us a step closer to generalizable autonomous driving.