Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

TrAISformer -- A Transformer Network with Sparse Augmented Data Representation and Cross Entropy Loss for AIS-based Vessel Trajectory Prediction

About

Vessel trajectory prediction plays a pivotal role in numerous maritime applications and services. While the Automatic Identification System (AIS) offers a rich source of information to address this task, forecasting vessel trajectory using AIS data remains challenging, even for modern machine learning techniques, because of the inherent heterogeneous and multimodal nature of motion data. In this paper, we propose a novel approach to tackle these challenges. We introduce a discrete, high-dimensional representation of AIS data and a new loss function designed to explicitly address heterogeneity and multimodality. The proposed model-referred to as TrAISformer-is a modified transformer network that extracts long-term temporal patterns in AIS vessel trajectories in the proposed enriched space to forecast the positions of vessels several hours ahead. We report experimental results on real, publicly available AIS data. TrAISformer significantly outperforms state-of-the-art methods, with an average prediction performance below 10 nautical miles up to ~10 hours.

Duong Nguyen, Ronan Fablet• 2021

Related benchmarks

TaskDatasetResultRank
Vessel Trajectory PredictionMaritime-MmD+ 1.0 (test)
ADE1.63
42
Vessel Trajectory PredictionMaritime-MmD+ v1 (test)
Average Displacement Error (ADE)2.4
42
Trajectory PredictionMaritime-MmD+ Prediction Step scenario
ADE ($10^{-2}$)212
28
Trajectory PredictionMaritime-MmD+ Vessel Density scenario
ADE385
28
Trajectory PredictionMaritime-MmD+ AIS Missing scenario
ADE (10^-2)2.58
28
Vessel Trajectory PredictionMaritime-MmD+ medium vessel density (Φ = ϕ_m)
ADE1.7
14
Vessel Trajectory PredictionMaritime-MmD+ low vessel density (Φ = ϕ_l)
ADE6.09
14
Vessel Trajectory PredictionMaritime-MmD+ high vessel density (Φ = ϕ_h)
ADE2.19
14
Vessel Destination EstimationWorldwide AIS Trajectory (test)
Overall Accuracy64.6
13
Vessel Trajectory PredictionPrivate Vessel Trajectory Dataset (test)
MSEP0.71
3
Showing 10 of 11 rows

Other info

Follow for update