mlsanityData to 14 Oct 2026

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

arXiv 2109.03958 ↗First posted 8 Sep 2021cs.AIcited by 20

Not on the map. Why?

Abstract (arXiv)

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.

Key facts. Cited by 20 papers in the map (data through 14 Oct 2026). Builds on 23 of its 78 references that are in the map. First posted on arXiv 8 Sep 2021.

Builds on

23 of 78 references.

References · 23

  1. 2017Attention Is All You Needcited by 60,361
  2. 2016SGDR: Stochastic Gradient Descent with Warm Restartscited by 5,045
  3. 2013Auto-Encoding Variational Bayescited by 4,053
  4. 2015Variational Inference with Normalizing Flowscited by 2,170
  5. 2020Knowledge Distillation: A Surveycited by 1,603
  6. 2019Explainable Artificial Intelligence (XAI): Concepts, Taxonomies, Opportunities and Challenges toward Responsible AIcited by 1,552
  7. 2018A Convergence Theory for Deep Learning via Over-Parameterizationcited by 787
  8. 2018Social GAN: Socially Acceptable Trajectories with Generative Adversarial Networkscited by 690
  9. 2020NVAE: A Deep Hierarchical Variational Autoencodercited by 637
  10. 2019Importance Estimation for Neural Network Pruningcited by 541
  11. 2015A Recurrent Latent Variable Model for Sequential Datacited by 362
  12. 2020TNT: Target-driveN Trajectory Predictioncited by 347
11 more
  1. 2018SoPhie: An Attentive GAN for Predicting Paths Compliant to Social and Physical Constraintscited by 298
  2. 2021DenseTNT: End-to-end Trajectory Prediction from Dense Goal Setscited by 295
  3. 2018On the Optimization of Deep Networks: Implicit Acceleration by Overparameterizationcited by 284
  4. 2019Human Motion Trajectory Prediction: A Surveycited by 232
  5. 2017Social Attention: Modeling Attention in Human Crowdscited by 186
  6. 2016Variational Autoencoder for Deep Learning of Images, Labels and Captionscited by 137
  7. 2021Deep Learning Methods for Vessel Trajectory Prediction based on Recurrent Neural Networkscited by 32
  8. 2019GeoTrackNet-A Maritime Anomaly Detector using Probabilistic Neural Network Representation of AIS Tracks and A Contrario Detectioncited by 16
  9. 2018A Multi-task Deep Learning Architecture for Maritime Surveillance using AIS Data Streamscited by 13
  10. 2020Space-based Global Maritime Surveillance. Part II: Artificial Intelligence and Data Fusion Techniquescited by 6
  11. 2020Detection of Abnormal Vessel Behaviours from AIS data using GeoTrackNet: from the Laboratory to the Oceancited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 1; 2024: 2; 2025: 9; 2026: 8.

20 papers.

Sort:

By area:

  1. 2023Multi-Path Long-Term Vessel Trajectories Forecasting with Probabilistic Feature Fusion for Problem Shiftingcited by 12
  2. 2025AIS Data-Driven Maritime Monitoring Based on Transformer: A Comprehensive Reviewcited by 6
  3. 2025Tensor-Var: Efficient Four-Dimensional Variational Data Assimilationcited by 4
  4. 2025AIS-LLM: A Unified Framework for Maritime Trajectory Prediction, Anomaly Detection, and Collision Risk Assessment with Explainable Forecastingcited by 4
  5. 2024Enhancing Maritime Trajectory Forecasting via H3 Index and Causal Language Modelling (CLM)cited by 3
  6. 2025Physics-Informed Neural Networks for Vessel Trajectory Prediction: Learning Time-Discretized Kinematic Dynamics via Finite Differencescited by 3
  7. 2024Incorporating Navigation Context into Inland Vessel Trajectory Prediction: A Gaussian Mixture Model and Transformer Approachcited by 2
  8. 2025Goal-Conditioned Reinforcement Learning for Data-Driven Maritime Navigationcited by 2
  9. 2026SKETCH: Semantic Key-Point Conditioning for Long-Horizon Vessel Trajectory Predictioncited by 1
  10. 2026CmIVTP: Cross-modal Interaction-based Vessel Trajectory Prediction for Maritime Intelligencecited by 1
  11. 2025Multi-vessel Interaction-Aware Trajectory Prediction and Collision Risk Assessmentcited by 0
  12. 2025An Automated Tip-and-Cue Framework for Optimized Satellite Tasking and Visual Intelligencecited by 0
8 more
  1. 2025Predictive Modeling of Maritime Radar Data Using Transformer Architecturecited by 0
  2. 2025Spatio-Temporal Graphs Beyond Grids: Benchmark for Maritime Anomaly Detectioncited by 0
  3. 2026Hierarchical Two-Stage Framework for Environment-Aware Long-Horizon Vessel Trajectory Predictioncited by 0
  4. 2026M-CTX: Exact and Scalable Spatial Context Retrieval for Trajectory Analyticscited by 0
  5. 2026NaviAIS: A Scenario-Level Vessel Trajectory Prediction Dataset withVectorized Lane Priors and the NaviLane Forecasting Frameworkcited by 0
  6. 2026Does Latent Context Help? A Controlled Evaluation of Inverse Reinforcement Learning in Arctic Shippingcited by 0
  7. 2026Non-Parametric Spatiotemporal Trajectory Prediction via State-Conditioned Transition Samplingcited by 0
  8. 2026M3-Former: Multimodal Transformer with Mixture-of-Experts for Long-Term Vessel Trajectory Predictioncited by 0