Research Specialist Associate - Deep Learning @CORL
I am an M.S. in Robotics from the University of Michigan, Ann Arbor. I’m a coding fanatic and hope to develop and implement techniques in my future endeavors that will improve the decision making capabilities of computer systems, hence bridging the communication gap between humans and machines.
Skills
I am a programming fanatic and have a solid foundation of robot programming as well as machine learning and deep learning frameworks
Resume
Education
Master of Science & Robotics
August 2021 - December 2022
University of Michigan, Ann Arbor, MI
- Cumulative GPA: 3.7
- Research: Heterogenous Mobile Manipulator Collaborating for On-Demand Sensing
- Relevant Coursework: Machine Learning, Deep Learning for Computer Vision, Mobile Robotics, Mathematics for Robotics, Robotics Systems Lab, Intermediate Dynamics
- Electives: Interpersonal Skills
Bachelor of Technology & Mechanical Engineering
July 2016 - May 2020
Jamia Millia Islamia, New Delhi, India
- Cumulative GPA:
4.0 - Graduated with First Divison with Honors
Professional Experience
Deep Learning Scientist
May 2022 - Present
Comparative Orthopaedic Rehabilitation Laboratory
- Automated the frame extraction process from training videos and incorporated them to train a transformer-based deep network to map the hip, knee, and ankle joints.
- Developed post processing script for peak detection and normalization of knee-flexion angle from the transformer network output.
- Integrated the workflow to connect Data Acquisition Unit (DAQ) with an amplifier and developed signal processing scripts for analyzing electromyographic signal data to map the network inferred data.
- Scripting FDA using PCA analysis for the GAIT data and prediction of time series data with machine learning models.
- Quantification of muscle fibers using UNET model
Graduate Researcher
January 2022 - April 2022
University of Michigan, Ann Arbor, MI
- Implemented the code to collect dataset for different orientations of the KUKA iiwa7 manipulator with Blender.
- Built an auto-encoder network to process the KUKA dataset and produce 7 belief maps, one each for every joint (key points).
- Combined the resultant 2D belief maps with the known camera intrinsics of the Toyota HSR robot and forward kinematics of KUKA using OpenCV PnP to get the relative pose of the base of KUKA w.r.t HSR frame.
Graduate Instructor
October 2021 - March 2022
Web Developer
June 2015 - December 2019
Sanaz Technologies, India
- Developed websites in HTML/CSS/Bootstrap, Django, and WordPress
- Reviewed websites and producing reports for their website optimization
- Integrated on-page, technical and off-page seo techniques to increase the ranking of the website on Google