Representing Spatial Trajectories as Distributions
Abstract
We introduce a representation learning framework for spatial trajectories. We represent partial observations of trajectories as probability distributions in a learned latent space, which characterize the uncertainty about unobserved parts of the trajectory. Our framework allows us to obtain samples from a trajectory for any continuous point in time—both interpolating and extrapolating. Our flexible approach supports directly modifying specific attributes of a trajectory, such as its pace, as well as combining different partial observations into single representations. Experiments show our method's superiority over baselines in prediction tasks.
Cite
Text
Coll-Vinent and Vondrick. "Representing Spatial Trajectories as Distributions." Neural Information Processing Systems, 2022.Markdown
[Coll-Vinent and Vondrick. "Representing Spatial Trajectories as Distributions." Neural Information Processing Systems, 2022.](https://mlanthology.org/neurips/2022/collvinent2022neurips-representing/)BibTeX
@inproceedings{collvinent2022neurips-representing,
title = {{Representing Spatial Trajectories as Distributions}},
author = {Coll-Vinent, Didac Suris and Vondrick, Carl},
booktitle = {Neural Information Processing Systems},
year = {2022},
url = {https://mlanthology.org/neurips/2022/collvinent2022neurips-representing/}
}