Hello! I'm Kshitij
I'm a ardent research professional with two years of professional experience in the field of robotics, artificial intelligence (AI) using deep learning - computer vision and reinforcement learning (RL). I got started with machine learning (ML) and AI with toy project of optical character recognition using deep learning. I was exposed to sophisticated ML and robotics through my Master's Degree in Robotics at Rochester Institute of Technology, New York. Working in industry for almost two years, I have had an opportunity to blend my experience and knowledge in creating solutions for real world problems applied through robots at production scale.
Kshitij Bichave
Deep Learning Nerd
EXPERIENCE
January 2020-Present
Deep Reinforcement Learning Engineer
iROBOT CORPORATION
Building cool stuff using Deep Reinforcement Learning.
September 2019-December 2019
Computer Vision Engineer Intern
PATH ROBOTICS
Worked in the R&D Computer Vision team at Path Robotics where I was successful in establishing new metrics for bench marking performance of the deep learning computer model aligning with product critical process and transformed the python process to C++. Also worked on trajectory planning in 3D space using computer vision and ROS.
January 2019-August 2019
Deep Learning Software Engineer Intern
iROBOT CORPORATION
Worked in R&D team to address problem statement using state of the art Reinforcement Learning algorithms, coupled with Sim2Real to transfer the model trained on physics engine (MuJoco) to real robot. Involved in environment setup, environment parameter tuning, training model with randomization to actual delivery on real world robot.
May 2018-December 2018
Graduate Research Assistant
ROCHESTER INSTITUTE OF TECHNOLOGY
Led a team of three to develop tensor decomposition-based algorithms to fuse modalities of the signals generated by MEMS for human motion recognition and human injury prediction using machine learning. Developed a multi-modal motion recognition for healthcare and safety applications using MEMS sensor data with tensor methods and machine learning/ deep learning.
EDUCATION
2017-2019
Master's Degree
ROCHESTER INSTITUTE OF TECHNOLOGY
Graduated with Master's in Electrical Engineering focused in Robotics. Completed an extensive coursework including Deep Learning, Image Processing and Computer Vision, Bio-Robotics, Deep 3D Scene Flow Estimation for SLAM. Worked with Dr. Panos Markopolous, Dr. Christopher Kanan and Dr. Guoyu Lu.
2013-2017
Bachelor's Degree
UNIVERSITY OF PUNE
Graduated with Bachelors of Engineering in Electronics and Telecommunication from Maharashtra Institute of Technology, Pune. Highlights of the coursework include - signal processing, electronic hardware circuit design, artificial intelligence. Worked heavily under the guidance of Dr. Alwin Anuse.
RECENT ACTIVITIES
April 2021
June 2020
INTRODUCTION TO RL PODCAST
​
August 2019
INTRODUCTION TO RL HANDS-ON WORKSHOP
Conducted an Introduction to Reinforcement Learning workshop at MIT WPU (Maharashtra Institute of Technology, Kothrud, Pune) last week. It was an honor to be back to alma mater, to conduct the workshop in the same room, where I once heard people talk about cool stuff. It was an incredible opportunity, to share my knowledge and skills in deep learning and reinforcement learning with your current students.
May 2021
April 2021
ICLR 2021- RL Highlights
INVITED TALK: AI AND ROBOTICS
Talked about advancements in AI and Robotics with examples of state of the art methods in computer vision and reinforcement learning. Attendees were the students of Mechanical Engineering and Electrical Engineering.
June 2020
INTRODUCTION TO RL PODCAST
​
August 2019
INTRODUCTION TO RL HANDS-ON WORKSHOP
Conducted an Introduction to Reinforcement Learning workshop at MIT WPU (Maharashtra Institute of Technology, Kothrud, Pune) last week. It was an honor to be back to alma mater, to conduct the workshop in the same room, where I once heard people talk about cool stuff. It was an incredible opportunity, to share my knowledge and skills in deep learning and reinforcement learning with your current students.
October 2018
Gait recognition based on tensor analysis of acceleration data from wearable sensors published at IEEE Western New York Image and Signal Processing Workshop, Rochester, NY
SKILLS
Tensorflow, PyTorch, Keras
Ray, Rllib, Tune
Python, Matlab, C++
Open - mindedness, Adaptability
Teamwork, Creativity
Problem Solving, Critical Thinking
