Joint Intent & Trajectory Prediction (MTITP GAN)
Multi-task conditional generative network (MTITP-WGAN) jointly classifying vessel intent and forecasting future multi-modal trajectory distributions (Chapter 7, Ph.D. Dissertation).
May 1, 2026
Presented two papers at IEEE CASE 2025
Presented work on early intent classification and proactive maritime threat prediction at the IEEE Conference on Automation Science and Engineering.
Aug 25, 2025
Feature-aware deep learning paper forthcoming at IEEE FMLDS 2025
Paper on feature-aware maritime intent recognition accepted for the IEEE International Conference on Future Machine Learning and Data Science.
Jan 15, 2025
NavySim: A Multi-Vessel Simulation and Analysis Engine for Naval Domains
We introduce a Unity-based naval simulator for maritime training that supports complex multi-vessel scenarios, real-time threat assessment, and intent recognition. The system models vessel capabilities, renders vulnerability and CPA-based action heatmaps, fuses them into a dynamic threat map, and uses HMMs to infer other agents’ intentions.
Aug 28, 2024
Intent Recognition for Maritime Autonomy
Deep and probabilistic intent classifiers (HMMs, LSTMs, Transformers) for early vessel behavior prediction in adversarial maritime scenarios.
Jun 1, 2022