Learning Agile Intruder Interception using Differentiable Quadrotor Dynamics
Michael Anoruo,
Xiaoyu Tian,
Abhishek Rathod,
Timothy Naudet,
Thomas Canchola,
Eric Sturzinger,
Kshitij Goel,
Wennie Tabib
arXiv preprint, 2026
We learn a control policy that intercepts an intruder using only 3D directional unit vectors from a passive monocular camera — without assuming access to relative position or distance. Using analytical policy gradients through differentiable quadrotor dynamics, our approach achieves agile interception at speeds up to 10 m/s, roughly a 30% improvement over point-mass baselines.
@article{anoruo2026,
title = {Learning Agile Intruder Interception using Differentiable Quadrotor Dynamics},
author = {Michael Anoruo and Xiaoyu Tian and Abhishek Rathod and Timothy Naudet and Thomas Canchola and Eric Sturzinger and Kshitij Goel and Wennie Tabib},
journal = {arXiv preprint},
year = {2026}
}
2025 1 paper
Automatic detection of cognitive events using machine learning and understanding models' interpretations of human cognition
Quang Dang,
Murat Kucukosmanoglu, Michael Anoruo,
Golshan Kargosha,
Sarah Conklin,
Justin Brooks
Automatic detection of cognitive events using machine learning and understanding models' interpretations of human cognition
Quang Dang,
Murat Kucukosmanoglu, Michael Anoruo,
Golshan Kargosha,
Sarah Conklin,
Justin Brooks
Scientific Reports, 2025
We train machine learning models on task-evoked pupillary response (pupil diameter and gaze position) to detect cognitive events across four domains — vigilance, emotion processing, numerical reasoning, and short-term memory — and examine how generalized models trade specificity for sensitivity relative to specialized ones.
@article{dang2025,
title = {Automatic detection of cognitive events using machine learning and understanding models' interpretations of human cognition},
author = {Quang Dang and Murat Kucukosmanoglu and Michael Anoruo and Golshan Kargosha and Sarah Conklin and Justin Brooks},
journal = {Scientific Reports},
year = {2025}
}
2022 1 paper
Knowledge Guided Two-Player Reinforcement Learning for Cyber Attacks and Defenses
Aritran Piplai, Michael Anoruo,
Kayode Fasaye,
Anupam Joshi,
Tim Finin,
Ahmad Ridley
IEEE International Conference on Machine Learning and Applications (ICMLA), 2022
Knowledge Guided Two-Player Reinforcement Learning for Cyber Attacks and Defenses
Aritran Piplai, Michael Anoruo,
Kayode Fasaye,
Anupam Joshi,
Tim Finin,
Ahmad Ridley
IEEE International Conference on Machine Learning and Applications (ICMLA), 2022
We present a two-player game environment in which attacker and defender agents are trained simultaneously in a simulated cyber battle. By incorporating guidance from Cybersecurity Knowledge Graphs on attack and mitigation steps, we accelerate agent convergence and improve defenses against previously unknown exploits.
@article{piplai2022,
title = {Knowledge Guided Two-Player Reinforcement Learning for Cyber Attacks and Defenses},
author = {Aritran Piplai and Michael Anoruo and Kayode Fasaye and Anupam Joshi and Tim Finin and Ahmad Ridley},
journal = {IEEE International Conference on Machine Learning and Applications (ICMLA)},
year = {2022}
}
