Searching-Eye
Contact:
  • sanjeev[dot]sharma[dot]iitr[AT]gmail[Dot]com
  • sanjeev1[AT]ualberta[dot]ca
  • sanjeev[At]searching-eye[dot]com
SANJEEV SHARMA (Dreams have a meaning, only if you care about them)!
.................................................................................................................................................................................................
I am a 2nd year masters student in the Computing Science department at the University of Alberta, Canada. I received my B.Tech in Electrical Engineering from the Indian Institute of Technology, Roorkee, in 2011.

Searching-Eye
During the second year of undergraduate (May 2009), I co-founded Searching-Eye with Prateek Gianchandani; I have delivered more than 35 lectures in machine learning, convex optimization and reinforcement learning through searching-eye.

Research
My research spans many areas of AI and Robotics. I am particularly interested in motion and path planning, autonomous navigation, robust control, agents coordination and convex optimization techniques for guaranteeing robot's safety and convergence to the goal during navigation in unknown environments. I am also interested in using constrained MDPs and POMDPs for bounded-risk path planning in uncertrain, dyanmic and unknown environments.

Inspiration
My research was inspired by the inverted helicopter project, and the 2007 DARPA Urban Challenge, especially by this video by the MIT. Working on such projects (developing such robots) is my primary objective.

Recently Delivered on Searching-Eye:
  • Unconstrained Minimization: Steepest Descent Methods & Convergence Analysis. (Jan 7th, 2011)
  • Unconstrained Minimization: Convergence Analysis of Gradient Descent Using Line Search. (12th Dec 2010)



Complete list of Lectures (delivered while an undergrad at IIT Roorkee)
Research Keywords:
  • Autonomous Navigation
  • Robotics
  • Motion and Path Planning
  • Reinforcement Learning
  • Convex Optimization
  • Machine Learning
  • Multi-Agent Systems
Education:
  • Master of Science (In Progress)
    Computing Science, University of Alberta
    2012 - Present; Current GPA: 3.90/4.0


    Effects of Time-Cycle in Reinforcement Learning Robotics
    (In Progress)

  • B. Tech, Electrical Engineering
    Indian Institute of Technology, Roorkee
    2007 - 2011;  GPA: 6.275/10


    Intrusion Detection Using Reinforcement Learning
    (Sanjeev Sharma and Prateek Gianchandani)
    A: 9/10

Important Links:
  • OMPL: The open motino planning library
    Great work; includes state of the art motion planners


        Instructions for installing the matplotlib for
        Python 3.2 are here
Research Interns:
  • Arial University, Israel
    - Paslin Laboratory of Robotics & Autonomous Vehicles: Under: Zvi Shiller
    - Developed an online motion planner
    - Publications: RoManSy 2012, IROS 2012, IJRR 2013
    - Sep 14, 2011 - April 20 2012
          IJRR Video

  • University of Massachusetts Amherst, USA
    Autonomous Learning Lab: Under: Sridhar Mahadevan
    - Developed the SPRPI algorithm: SPRPI
    - May 15 - Aug 05, 2010




Representative Research Projects (Self-Initiated):
  • DFRL-E Project:
    A coordinated RL agent based navigation framework for navigation in completely unknown and unseen 2D and 3D spaces with any existing path or motion planner.

  • ECAN Project:
    Convex optimization based path planner for planning in completely unknown and unseen environments (2D or 3D) using convex optimization and reinforcement learning.