Abstract:Trajectory data is essential for various applications as it records the movement of vehicles. However, publicly available trajectory datasets remain limited in scale due to privacy concerns, which hinders the development of trajectory data mining and trajectory-based applications. To address this issue, some methods for generating synthetic trajectories have been proposed to expand the scale of the dataset. However, all existing methods generate trajectories in the geographical coordinate system, which poses two limitations for their utilization in practical applications: 1) the inability to ensure that the generated trajectories are constrained on the road. 2) the lack of road-related information. In this paper, we propose a new problem to meet the practical application need, \emph{i.e.}, road network-constrained trajectory (RNTraj) generation, which can directly generate trajectories on the road network with road-related information. RNTraj is a hybrid type of data, in which each point is represented by a discrete road segment and a continuous moving rate. To generate RNTraj, we design a diffusion model called Diff-RNTraj. This model can effectively handle the hybrid RNTraj using a continuous diffusion framework by incorporating a pre-training strategy to embed hybrid RNTraj into continuous representations. During the sampling stage, a RNTraj decoder is designed to map the continuous representation generated by the diffusion model back to the hybrid RNTraj format. Furthermore, Diff-RNTraj introduces a novel loss function to enhance the spatial validity of the generated trajectories. Extensive experiments conducted on two real-world trajectory datasets demonstrate the effectiveness of the proposed model.

When Self-attention and Topological Structure Make a Difference: Trajectory Modeling in Road Networks.

Trajectory data-based traffic flow studies: A revisit

Visual Analysis of Traffic Data Based on Topic Modeling (chinavis 2017).

Trajectory-as-a-Sequence: A Novel Travel Mode Identification Framework

Trajectory Improves Data Delivery in Urban Vehicular Networks.

NetTraj: A Network-based Vehicle Trajectory Prediction Model with Directional Representation and Spatiotemporal Attention Mechanisms

ControlTraj: Controllable Trajectory Generation with Topology-Constrained Diffusion Model

TR2RM: an urban road network generation model based on multisource big data

Visual Analytics of Taxi Trajectory Data Via Topical Sub-trajectories.

Learning to Generate Maps from Trajectories

Diff-RNTraj: A Structure-aware Diffusion Model for Road Network-constrained Trajectory Generation

Characterized Diffusion and Spatial-Temporal Interaction Network for Trajectory Prediction in Autonomous Driving

Trajectory Topic Modelling to Characterize Driving Behaviors with GPS-Based Trajectory Data

MTrajRec

MFTraj: Map-Free, Behavior-Driven Trajectory Prediction for Autonomous Driving

Robust Dynamic Trajectory Regression on Road Networks: A Multi-task Learning Framework

Ego-centric Traffic Behavior Understanding Through Multi-Level Vehicle Trajectory Analysis.

Trajectory Prediction for Autonomous Driving with Topometric Map

Modeling Trajectories with Recurrent Neural Networks.

Characterized Diffusion Networks for Enhanced Autonomous Driving Trajectory Prediction

Label-Based Trajectory Clustering in Complex Road Networks