I am a doctoral student in the Computer Science department at the University of Maryland, College Park. I completed my Masters of Engineering in Robotics program from here in 2017. I am a member of the Perception and Robotics Group (PRG) pursuing research under the guidance of Professor Yiannis Aloimonos and Dr. Cornelia Fermüller in robot grasping, perception and cognitive robotics.
I received my Bachelor's degree in Computer Science and Engineering from the Birla Institute of Technology, Mesra in 2015.
Researching grasping using tactile feedback control using the Shadow Dexterous Hand and UR10 manipulator.
Worked with and developed for Baxter and Ridgeback robots. Experienced in ROS.
Worked on 3D vision for segmenting and detecting objects in scene, for grasping pipeline.
Using RGB-D cameras such as Kinect and XTion, and libraries like PCL, Open3D and cilantro.
Developed appliance detection and recognition pipeline for the Baxter.
Developed web-based user interface using ROS and JS for controlling Baxter. Project done as part of summer internship.
Developed for the Hololens and Oculus Rift for various projects, including human-robot collaboration interfaces.
Researched reinforcement learning algorithms for UR10 manipulator planning and execution, with dynamic obstacle avoidance. Created custom Gazebo-based simulation environment with OpenAI Gym bindings, and a voxel grid representation suitable for the network architecture. Developed using C++, Python and ROS.
Developed machine learning software for automated presentation analysis and evaluation. Multi-modal approach, using separate networks for speech, audio and facial feature analysis. Also presented options for cloud deployment using IBM and Microsoft APIs. Used Python, scikit-learn, OpenCV and dlib
Startup working on developing a home assistance robot system.
Worked with Baxter and Ridgeback robots.
Developed web-based ROS integrated user interface to control robot through mobile devices. Used ROS Web Tools, ROSLibJS and HTML5/jQuery for interface and Wit.AI and BASH for speech and NLP.
Developed ROS service for Ridgeback movement through PID implementation in Python.
Developed microwave LCD visual verification system using OpenCV and Python.
Kanishka Ganguly — kganguly@umd.edu