Averaging GPS segments competition 2019
作者：
Highlights：
• We organized a challenge where several participants attempted to find the best trajectory averaging method.
• We implemented a new baseline using components received from the submissions. The baseline is fast and outperforms the submitted methods.
• The choice of sequence averaging method is important.
• Removing outliers using several methods has an over-fitting effect. In general results are worse on new data.
摘要
•We organized a challenge where several participants attempted to find the best trajectory averaging method.•We implemented a new baseline using components received from the submissions. The baseline is fast and outperforms the submitted methods.•The choice of sequence averaging method is important.•Removing outliers using several methods has an over-fitting effect. In general results are worse on new data.
论文关键词：Gps trajectories,Trajectory averaging,Time series,Representative segment
论文评审过程：Received 11 July 2019, Revised 24 August 2020, Accepted 29 October 2020, Available online 30 October 2020, Version of Record 30 January 2021.
论文官网地址：https://doi.org/10.1016/j.patcog.2020.107730
