![[2602.13071] Bus-Conditioned Zero-Shot Trajectory Generation via Task Arithmetic](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.13071] Bus-Conditioned Zero-Shot Trajectory Generation via Task Arithmetic
Summary
This paper introduces MobTA, a novel approach for generating mobility trajectories without requiring real data from the target city, using bus timetables and source city data.
Why It Matters
The research addresses the challenge of generating mobility trajectories in scenarios where data is scarce, which is crucial for advancing smart city applications. By leveraging existing bus timetables and data from other cities, this method enhances the applicability of trajectory generation techniques, potentially improving urban planning and transportation systems.
Key Takeaways
- MobTA enables trajectory generation without real mobility data from the target city.
- The method utilizes bus timetables and source city data for effective modeling.
- Task arithmetic is introduced to enhance trajectory generation accuracy.
- MobTA shows significant performance improvements over existing methods.
- The approach is theoretically analyzed for stability across different models.
Computer Science > Machine Learning arXiv:2602.13071 (cs) [Submitted on 13 Feb 2026] Title:Bus-Conditioned Zero-Shot Trajectory Generation via Task Arithmetic Authors:Shuai Liu, Ning Cao, Yile Chen, Yue Jiang, Gao Cong View a PDF of the paper titled Bus-Conditioned Zero-Shot Trajectory Generation via Task Arithmetic, by Shuai Liu and 4 other authors View PDF HTML (experimental) Abstract:Mobility trajectory data provide essential support for smart city applications. However, such data are often difficult to obtain. Meanwhile, most existing trajectory generation methods implicitly assume that at least a subset of real mobility data from target city is available, which limits their applicability in data-inaccessible scenarios. In this work, we propose a new problem setting, called bus-conditioned zero-shot trajectory generation, where no mobility trajectories from a target city are accessible. The generation process relies solely on source city mobility data and publicly available bus timetables from both cities. Under this setting, we propose MobTA, the first approach to introduce task arithmetic into trajectory generation. MobTA models the parameter shift from bus-timetable-based trajectory generation to mobility trajectory generation in source city, and applies this shift to target city through arithmetic operations on task vectors. This enables trajectory generation that reflects target-city mobility patterns without requiring any real mobility data from it. Furthermore, ...
