I am a PhD candidate in the Department of Computer Science and Engineering at the University of Minnesota, Twin Cities. My advisor is Nikolaos Papanikolopoulos and I'm a research assistant at the Center for Distributed Robotics.
Email: beksi (at) cs.umn.edu
4-192 Keller Hall
200 Union Street SE
Minneapolis, MN 55455
My research focuses on developing algorithms and data structures for fundamental robotic perception tasks such as 3D reconstruction, segmentation, and object detection and classification. I investigate the use of applied topology, specifically topological data analysis for processing massive sensor datasets. I also work on problems in networked and cloud robotics where the goal is to allow under-resourced robots remote access to computing, data, and learning resources. A portfolio of the software I've developed for my research can be found on GitHub.
- I will be showing my work on the parallel construction of the Vietoris-Rips complex in 3D at the Minnesota Supercomputing Institute Research Exhibition.
- I have been awarded the MnDRIVE PhD Fellowship through the University of Minnesota Informatics Institute for my research on topological methods for 3D point cloud processing.
- This summer I will be working as a Robotics Engineer in the Technology Organization at iRobot. I'll be spearheading a project for synthetic 3D scene generation. The goal of this project is to generate synthetic 3D scenes suitable for training data-driven large-scale machine learning methods such as deep learning.
- I'll be traveling to Daejeon, South Korea, to talk about my research on 3D region segmentation using topological persistence at IROS 2016.
- I will present my research on 3D point cloud segmentation using topological persistence at ICRA 2016 in Stockholm, Sweden.
- I've been invited to attend the ICRA 2016 Workshop on Emerging Topological Techniques in Robotics and present a poster highlighting my recent work.
- Our work on covariance based point cloud descriptors for object detection and recognition has been published in the journal Computer Vision and Image Understanding.
- The Cloud-based Object Recognition Engine (CORE) we are developing has been accepted for publication at IROS 2015 in Hamburg, Germany.
- I will present my research on a dictionary learning framework using RGB-D covariance descriptors on point cloud data for performing object classification at ICRA 2015 in Seattle, Washington.
- I've been invited to attend the NSFCloud Workshop on Experimental Support for Cloud Computing to provide input on experimental requirements for cloud robotics.
- My research on point cloud culling for robot vision tasks has been selected for presentation at IROS 2014 in Chicago, Illinois.
- I will be at ICRA 2014 in Hong Kong to present our work on object classification using RGB-D covariance descriptors along with an implementation of this work on a mobile robot.
- I have been invited to give an oral presentation at ICPEAL 2012 in Nagoya, Japan, regarding my work on Chinese word segmentation. This work is a result of my project on natural language processing (NLP) for the Artificial Intelligence class (CSCI 5511) at the University of Minnesota.
The views and opinions expressed in this page are strictly those of the page author.
The contents of this page have not been reviewed or approved by the University of Minnesota.