arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2103.16101cs.CV
收藏

基于自监督特征提取的大规模自动驾驶场景聚类

Large Scale Autonomous Driving Scenarios Clustering with Self-supervised Feature Extraction

Baidu Research(百度研究院);National Engineering Laboratory of Deep Learning Technology and Application(深度学习技术及应用国家工程实验室)

Jinxin Zhao, Jin Fang, Zhixian Ye, Liangjun Zhang

摘要:

自动驾驶场景数据的聚类可以通过提高仿真测试的完整性和逼真度来大大有益于自动驾驶验证和仿真系统。本文提出了一个全面的数据聚类框架,为大量的车辆驾驶数据。现有的算法利用手工制作的功能,其质量依赖于人类专家的判断。此外,相关的特征压缩方法对于大数据集是不可缩放的。我们的方法充分考虑了交通元素,包括交通代理对象和地图信息。同时,我们提出了一种用于空间和时间特征提取的自监督深度学习方法,以避免有偏见的数据表示。新设计的基于数据增强的驾驶数据聚类评价指标,准确性评估不需要人为标记的数据集,这是受人的偏见。通过这种公正的评估指标,我们已经表明,我们的方法超越了现有的方法,依赖于手工特征提取。

英文摘要:

The clustering of autonomous driving scenario data can substantially benefit the autonomous driving validation and simulation systems by improving the simulation tests' completeness and fidelity. This article proposes a comprehensive data clustering framework for a large set of vehicle driving data. Existing algorithms utilize handcrafted features whose quality relies on the judgments of human experts. Additionally, the related feature compression methods are not scalable for a large data-set. Our approach thoroughly considers the traffic elements, including both in-traffic agent objects and map information. Meanwhile, we proposed a self-supervised deep learning approach for spatial and temporal feature extraction to avoid biased data representation. With the newly designed driving data clustering evaluation metrics based on data-augmentation, the accuracy assessment does not require a human-labeled data-set, which is subject to human bias. Via such unprejudiced evaluation metrics, we have shown our approach surpasses the existing methods that rely on handcrafted feature extractions.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑