My current work focuses on 3D machine perception for AR/VR applications. More specifically, I've worked on mapping, object detection and tracking, with the goal of building and keeping up-to-date a semantic map of the world. This could enable a variety of applications, such as contextualized AI.Prior to that, my graduate study focused on the control and decision-making for robotic systems. In particular, for improving safety, I developed machine learning algorithms for predicting pedestrians'​ intents and behavior patterns. Furthermore, I worked on cooperative motion planning techniques for enabling autonomous navigation in complex indoor environments containing dynamic obstacles (i.e., pedestrians).