Abstract:With the development of location-acquisition technologies, there are a huge number of mobile trajectories generated and accumulated in a variety of domains. However, due to the constraints of device and environment, many trajectories are recorded at low sampling rate, which increases the uncertainty between two consecutive sampled points in the trajectories. Our task is to recover a high-sampled trajectory based on the irregular low-sampled trajectory in free space, i.e., without road network information. There are two major problems with traditional solutions. First, many of these methods rely on heuristic search algorithms or simple probabilistic models. They cannot well capture complex sequential dependencies or global data correlations. Second, for reducing the predictive complexity of the unconstrained numerical coordinates, most of the previous studies have adopted a common preprocessing strategy by mapping the space into discrete units. As a side effect, using discrete units is likely to bring noise or inaccurate information. Hence, a principled post-calibration step is required to produce accurate results, which has been seldom studied by existing methods. To address the above difficulties, we propose a novel Deep Hybrid Trajectory Recovery model, named DHTR. Our recovery model extends the classic sequence-to-sequence generation framework by implementing a subsequence-to-sequence recovery model tailored for the current task, named subseq2seq. In order to effectively capture spatiotemporal correlations, we adopt both spatial and temporal attentions for enhancing the model performance. With the attention mechanisms, our model is able to characterize long-range correlations among trajectory points. Furthermore, we integrate the subseq2seq with a calibration component of Kalman filter (KF) for reducing the predictive uncertainty. At each timestep, the noisy predictions from the subseq2seq component will be fed into the KF component for calibration, and then the refined predictions will be forwarded to the subseq2seq component for the computation of the next timestep. Extensive results on real-world datasets have shown the superiority of the proposed model in both performance and interpretability.

Arbitrary Spatial Trajectory Reconstruction based on A Single Inertial Sensor

CLINS: Continuous-Time Trajectory Estimation for LiDAR-Inertial System

Sensor-Fusion-Based Trajectory Reconstruction For Mobile Devices

An $\Omega$-Free Accelerometer Array Method for 2-D Trajectory Reconstruction

A Gaussian Mixture Unscented Rauch–Tung–Striebel Smoothing Framework for Trajectory Reconstruction

Gps-Simulated Trajectory Detection

An Omega-Free Gyro-Free Accelerometer Pair Algorithm For 2d Trajectory Reconstruction

Marine Trajectory Reconstruction Method Based on Navigation State Recognition and Bi-Directional Kinematic Interpolation

Indoor PDR Trajectory Matching by Gyroscope and Accelerometer Signal Sequence without Initial Motion State

SimiDTR: Deep Trajectory Recovery with Enhanced Trajectory Similarity

Traj-LIO: A Resilient Multi-LiDAR Multi-IMU State Estimator Through Sparse Gaussian Process

An <inline-formula><tex-math notation="LaTeX">$\omega$</tex-math></inline-formula>-Free Gyro-Free Accelerometer Pair Algorithm for 2D Trajectory Reconstruction

Real-Time 3D Motion Tracking and Reconstruction System Using Camera and IMU Sensors

3DTRIP: A General Framework for 3D Trajectory Recovery Integrated with Prediction.

Sensor-Fusion-Based Trajectory Reconstruction for Quadrotor Drones

High-resolution Reconstruction of Non-Motorized Trajectory in Shared Space: A New Approach Integrating the Social Force Model and Particle Filtering

Walking Trajectory Estimation Using Multi-Sensor Fusion and a Probabilistic Step Model

Deep Trajectory Recovery with Fine-Grained Calibration Using Kalman Filter

Towards Robust and Accurate Similar Trajectory Discovery: Weak-Parametric Approaches.

Robust Inertial Motion Tracking Through Deep Sensor Fusion Across Smart Earbuds and Smartphone.

Transformer Inertial Poser: Real-time Human Motion Reconstruction from Sparse IMUs with Simultaneous Terrain Generation