Search papers, labs, and topics across Lattice.
This study introduces a Deep Learning Enhanced Positioning, Velocity, and Time (PVT) algorithm designed to address multipath interference in densely populated urban environments. By employing a supervised objective to predict range corrections and uncertainty, alongside a Jointly Embedded Pretrained Algorithm (JEPA) for self-supervised learning, the researchers significantly enhance the algorithm's representation quality. The evaluation across various driving scenarios demonstrates a marked improvement in PVT accuracy, especially in challenging urban conditions, underscoring the effectiveness of leveraging unlabelled GNSS data for better generalization.
Unlabelled GNSS data can dramatically enhance PVT accuracy in urban environments, achieving significant improvements even under harsh conditions.
This work proposes a Deep Learning Enhanced PVT algorithm to mitigate multipath interference in dense urban areas. A supervised objective jointly predicts range corrections and uncertainty, while a JEPA-based self-supervised pretraining stage improves representation quality. The algorithm is evaluated over diverse driving scenarios, substantially improving PVT accuracy, particularly for unseen harsh urban conditions. These results highlight the potential of unlabelled GNSS data to improve generalization performance.
