GPS-MTM: Capturing Pattern of Normalcy in GPS-Trajectories with self-supervised learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Garg, Umang, Zhang, Bowen, Subrahmanya, Anantajit, Gudavalli, Chandrakanth, Manjunath, BS
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915537946148864
author Garg, Umang
Zhang, Bowen
Subrahmanya, Anantajit
Gudavalli, Chandrakanth
Manjunath, BS
author_facet Garg, Umang
Zhang, Bowen
Subrahmanya, Anantajit
Gudavalli, Chandrakanth
Manjunath, BS
contents Foundation models have driven remarkable progress in text, vision, and video understanding, and are now poised to unlock similar breakthroughs in trajectory modeling. We introduce the GPSMasked Trajectory Transformer (GPS-MTM), a foundation model for large-scale mobility data that captures patterns of normalcy in human movement. Unlike prior approaches that flatten trajectories into coordinate streams, GPS-MTM decomposes mobility into two complementary modalities: states (point-of-interest categories) and actions (agent transitions). Leveraging a bi-directional Transformer with a self-supervised masked modeling objective, the model reconstructs missing segments across modalities, enabling it to learn rich semantic correlations without manual labels. Across benchmark datasets, including Numosim-LA, Urban Anomalies, and Geolife, GPS-MTM consistently outperforms on downstream tasks such as trajectory infilling and next-stop prediction. Its advantages are most pronounced in dynamic tasks (inverse and forward dynamics), where contextual reasoning is critical. These results establish GPS-MTM as a robust foundation model for trajectory analytics, positioning mobility data as a first-class modality for large-scale representation learning. Code is released for further reference.
format Preprint
id arxiv_https___arxiv_org_abs_2509_24031
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle GPS-MTM: Capturing Pattern of Normalcy in GPS-Trajectories with self-supervised learning
Garg, Umang
Zhang, Bowen
Subrahmanya, Anantajit
Gudavalli, Chandrakanth
Manjunath, BS
Machine Learning
Artificial Intelligence
Computer Vision and Pattern Recognition
Multiagent Systems
Foundation models have driven remarkable progress in text, vision, and video understanding, and are now poised to unlock similar breakthroughs in trajectory modeling. We introduce the GPSMasked Trajectory Transformer (GPS-MTM), a foundation model for large-scale mobility data that captures patterns of normalcy in human movement. Unlike prior approaches that flatten trajectories into coordinate streams, GPS-MTM decomposes mobility into two complementary modalities: states (point-of-interest categories) and actions (agent transitions). Leveraging a bi-directional Transformer with a self-supervised masked modeling objective, the model reconstructs missing segments across modalities, enabling it to learn rich semantic correlations without manual labels. Across benchmark datasets, including Numosim-LA, Urban Anomalies, and Geolife, GPS-MTM consistently outperforms on downstream tasks such as trajectory infilling and next-stop prediction. Its advantages are most pronounced in dynamic tasks (inverse and forward dynamics), where contextual reasoning is critical. These results establish GPS-MTM as a robust foundation model for trajectory analytics, positioning mobility data as a first-class modality for large-scale representation learning. Code is released for further reference.
title GPS-MTM: Capturing Pattern of Normalcy in GPS-Trajectories with self-supervised learning
topic Machine Learning
Artificial Intelligence
Computer Vision and Pattern Recognition
Multiagent Systems
url https://arxiv.org/abs/2509.24031