CondTraj-GAN: Conditional Sequential GAN for Generating Synthetic Vehicle Trajectories

Nils Henke, Shimon Wonsak, Prasenjit Mitra, Michael Nolting, Nicolas Tempelmeier

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

While the ever-increasing amount of available data has enabled complex machine learning algorithms in various application areas, maintaining data privacy has become more and more critical. This is especially true for mobility data. In nearly all cases, mobility data is personal and therefore the drivers’ privacy needs to be protected. However, mobility data is particularly hard to anonymize, hindering its use in machine learning algorithms to its full potential. In this paper, we address these challenges by generating synthetic vehicle trajectories that are not subject to personal data protection but have the same statistical characteristics as the originals. We present CondTraj-GAN– Conditional Trajectory Generative Adversarial Network. – a novel end-to-end framework to generate entirely synthetic vehicle trajectories. We introduce a specialized training and inference procedure that enables the application of GANs to discrete trajectory data conditioned on their sequence length. We demonstrate the data utility of the synthetic trajectories by comparing their spatial characteristics with the original dataset. Finally, our evaluation shows that CondTraj-GAN reliably outperforms state-of-the-art trajectory generation baselines.

Original languageEnglish (US)
Title of host publicationAdvances in Knowledge Discovery and Data Mining - 27th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2023, Proceedings
EditorsHisashi Kashima, Tsuyoshi Ide, Wen-Chih Peng
PublisherSpringer Science and Business Media Deutschland GmbH
Pages79-91
Number of pages13
ISBN (Print)9783031333767
DOIs
StatePublished - 2023
Event27th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2023 - Osaka, Japan
Duration: May 25 2023May 28 2023

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume13936 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference27th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2023
Country/TerritoryJapan
CityOsaka
Period5/25/235/28/23

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • General Computer Science

Cite this