Shayegan Omidshafiei

  shayegan [at] mit [dot] edu
  Google Scholar
  GitHub

I am a Ph.D. candidate at the Laboratory for Information and Decision Systems (LIDS) and Aerospace Controls Lab (ACL) at MIT, advised by Prof. Jonathan How. Previously, I received my B.A.Sc. in Engineering Science at the University of Toronto.

I am interested in multiagent coordination, reinforcement learning, hierarchical learning, game theory, and augmented reality applications. I am a strong believer that the intelligent agents of tomorrow must not only be proficient learners, but also capable collaborators and communicators.

Updates

01/2018 Crossmodal Attentive Skill Learner was accepted to AAMAS 2018 as a full talk!
12/2017 Our paper Crossmodal Attentive Skill Learner was presented at NIPS 2017 as an oral talk at the Hierarchical Reinforcement Learning workshop, and as a poster at the Deep Reinforcement Learning Symposium.
10/2017 I passed my Ph.D. thesis proposal defense!
08/2017 Our paper on Multitask Multiagent Reinforcement Learning was presented at ICML 2017.
06/2017 I was proud to co-organize the Robot Communication in the Wild workshop at RSS 2017, held at MIT. Many fantastic presenters!
05/2017 Our 2 papers on decentralized decision-making using macro-observations and continuous observations were presented at ICRA 2017.
03/2017 Our paper integrating macro-actions into Dec-POMDP planning was accepted to IJRR!
11/2016 Our paper on augmented reality for rapid prototyping of robotics hardware was accepted into Control Systems Magazine, as cover article! Check out some videos of it in action here, here, and here!
05/2016 Our paper on graph-based multiagent decision making with macro-actions was presented at ICRA 2016!
08/2015 Wrapped up a internship at Qualcomm, working on deep learning-based perception for aerial robots.
05/2015 Our work introducing the Decentralized Partially Observable Semi-Markov Decision Process (Dec-POSMDP) was presented at ICRA 2015!

Research

Below is a selection of my publications. For full list, please visit my Google Scholar page.

Multiagent Planning & Learning
Single-agent Planning & Learning
Augmented Reality

Courses

Below is a selection of courses I have taken at MIT and University of Toronto.

AI, Controls, Computer Science
Principals of Autonomy & Decision Making
Algorithmic Game Theory & Data Science
Computational Cognitive Science
Machine Learning
Systems Optimization
Stochastic Estimation & Control
Principles of Optimal Control
Feedback Control Systems
Foundations of Planning Algorithms (audited)
Advances in Computer Vision (audited)
Spacecraft Dynamics & Control
Control Systems
Space Systems Design
Engineering Design
Scientific Computing
Digital & Computer Systems
Computer Programming & Data Structures
Math & Finance
Nonlinear Optimization
Probability & Statistics in Engineering
Probability & Statistics
Vector Calculus & Fluid Mechanics
Partial Differential Equations
Complex Analysis
Calculus III
Calculus II
Calculus I
Matrix & Vector Algebra
Linear Algebra
Mathematical Theory of Finance
Economic Analysis & Decision Making
Physics & Aero/Astro
Dynamics
Classical Mechanics
Aerodynamics
Aerospace Propulsion
Gasdynamics
Combustion Processes
Thermodynamics
Origin & Evolution of the Universe
Relativity
Quantum & Thermal Physics
Waves & Modern Physics
Fundamentals of Electric Circuits
Electricity & Magnetism
Molecules & Materials
Advanced Mechanics Of Structures
Mechanics of Solids & Structures
Structures & Materials

Press

Some of my works have been featured in technology news websites.