Editable Map-Conditioned Trajectory Generation for Human Mobility Simulation
Abstract
Geospatial simulation of infrastructure interventions requires mobility generators that respond directly to edited maps, yet many data-driven generators do not expose the map as an editable condition. We formulate this task as map-conditioned autoregressive generation of human mobility: a road raster conditions a decoder that emits nominal 31.25 m mesh-cell tokens at one-minute intervals. The mesh-local vocabulary supports held-out and locally edited maps without retraining or vocabulary changes. We instantiate a ResNet-50 visual-prefix configuration and a Vision Transformer (ViT) cross-attention configuration, trained from scratch on 87,400 smartphone-derived trajectories from 874 meshes in Ishikawa Prefecture, Japan; 219 meshes are held out. We evaluate map sensitivity by comparing correct-map and within-split shuffled-map generations with held-out real trajectories. On the 110-mesh test split, for the ResNet-50 configuration, correct-map generations are closer than shuffled-map generations on 60% of meshes under Hausdorff-based energy distance (p = 0.021), while DTW is directional but inconclusive (57%, p = 0.074); correlation with real density is 0.38 with the correct map versus 0.01 with shuffled maps. The ViT configuration shows weaker trajectory-level sensitivity and smaller density gains. An illustrative bridge-removal edit changes generated continuations without retraining. Together, these results support the feasibility of editable-map human-mobility simulation.
Versions
Related
- A Guidewire Trajectory Generation Tool for Interventional MRI Simulations
- Manufacturing-conditioned hierarchical generation of cathode active-material microstructures with global electrode context
- Atomic Design Transformer: Scaffold-Conditioned 3D Molecule Generation with xTB-Verified Reinforcement Learning
- Walk on Multiscale Media: Compiling Editable and Composable Transport Medium Prefabs for Monte Carlo PDE Solvers
- HarP: a simulation tool for high-order harmonic generation and propagation in gases
BibTeX
@misc{mizuno2026editable,
title = {{Editable Map-Conditioned Trajectory Generation for Human Mobility Simulation}},
author = {Takayuki Mizuno and Shouji Fujimoto and Mikito Hiruki and Atushi Ishikawa},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.32360},
eprint = {2609.32360},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.32360},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Editable Map-Conditioned Trajectory Generation for Human Mobility Simulation
AU - Takayuki Mizuno
AU - Shouji Fujimoto
AU - Mikito Hiruki
AU - Atushi Ishikawa
PY - 2026
DA - 2026/09/26
PB - arXiv
DO - 10.48550/arxiv.2609.32360
UR - https://arxiv.org/abs/2609.32360
AB - Geospatial simulation of infrastructure interventions requires mobility generators that respond directly to edited maps, yet many data-driven generators do not expose the map as an editable condition. We formulate this task as map-conditioned autoregressive generation of human mobility: a road raster conditions a decoder that emits nominal 31.25 m mesh-cell tokens at one-minute intervals. The mesh-local vocabulary supports held-out and locally edited maps without retraining or vocabulary changes. We instantiate a ResNet-50 visual-prefix configuration and a Vision Transformer (ViT) cross-attention configuration, trained from scratch on 87,400 smartphone-derived trajectories from 874 meshes in Ishikawa Prefecture, Japan; 219 meshes are held out. We evaluate map sensitivity by comparing correct-map and within-split shuffled-map generations with held-out real trajectories. On the 110-mesh test split, for the ResNet-50 configuration, correct-map generations are closer than shuffled-map generations on 60% of meshes under Hausdorff-based energy distance (p = 0.021), while DTW is directional but inconclusive (57%, p = 0.074); correlation with real density is 0.38 with the correct map versus 0.01 with shuffled maps. The ViT configuration shows weaker trajectory-level sensitivity and smaller density gains. An illustrative bridge-removal edit changes generated continuations without retraining. Together, these results support the feasibility of editable-map human-mobility simulation.
ER -