Selected projects in cardiovascular AI, autonomous systems, and generative sequence modeling.
Physics-consistent multi-agent naval simulator with custom shaders, real-time CPA threat heatmaps, and ML-driven intent prediction (2022–2026).
Fusing IVUS, OCT, and X-ray Angiography for coronary lesion analysis and agentic clinical decision support in percutaneous coronary interventions (PCI).
Multi-task conditional generative network (MTITP-WGAN) jointly classifying vessel intent and forecasting future multi-modal trajectory distributions (Chapter 7, Ph.D. Dissertation).
Systematic empirical evaluation of classical Bayesian filters and state estimators for multi-vessel trajectory prediction across horizons and noise regimes (Chapter 6, Ph.D. Dissertation).
User perception study evaluating embodied perspective-taking and torso-based role shifting on the Unitree G1 humanoid robot (Fall 2025).
Real-time heatmap framework fusing sensor coverage, vulnerability fields, and CPA-based threat estimates for naval situational awareness (2021–2023).
Earlier work in medical image analysis and neural network architectures.
Deep and probabilistic intent classifiers (HMMs, LSTMs, Transformers) for early vessel behavior prediction in adversarial maritime scenarios.
Graph Convolutional Neural Networks with Mask-RCNN backend for bilateral and ipsilateral mammogram fusion, malignancy detection, and semantic segmentation.
Segmentation pipelines and Local Haar Pattern descriptor for retinal fundus images, achieving state-of-the-art accuracy on DRIVE, STARE, and CHASE_DB1.