Rakesh Kumar

Research Engineer Control System and Deep Learning

OLA Electric

Biography

Hey! I am currently at OLA Electric working on 3D scene representation of end-to-end autonomous driving agents and self-supervised depth estimation and Mapping and Localization. On the development side of things, I have been involved in porting the PyTorch model to the TensorRT model for faster inference on resource-constraint devices. Developing ROS2 wrapper around the deep learning pipeline.

Previously, I have done my master’s from IISc Bangalore in Mechanical Engineering, Where I have worked on Optimal control in field of Thermal power plant.

Our work on robust pose graph SLAM has been published at the ICRA conference and our work on Robust feature matching has been presented at the IROS conference. These works targets improving SLAM performance in feature-less regions and improving correspondence matching in high illumination and viewpoint variations.

Download my resumé.

Interests

Optimal Control
Physics informed neural network
Optimization
Robotics

Education

M.Tech Course work in Mechanical Engineering, 2021
Indian Institute of science, bangalore
B.Tech in Mechanical, 2016
Indian Institute of Technology (ISM), Dhanbad

Experience

Computer Vision and SLAM Research Engineer

OLA Electric

Jul 2021 – Jan 2023 Bangalore

Developed an end to end autonomous driving agent using cameras, GPS and IMU sensors. Ported the agent from Carla simulator to NuScenes Dataset. Converted the pytorch model to TensorRT and developed a ROS wrapper to run on real Mahindra E2O car achieving final control prediction at 25 HZ, in a zero shot paradigm.
Extended the Lidar based mapping and localization LeGO-LOAM SLAM for the Velodyne and Ouster lidars and ported ROS1 to ROS2 in C++.
Won the silver medal in the Kaggle Image Matching Challenge 2022 by developing an Ensemble of Deep feature matching algorithm of SuperGlue and LoFTR.
Trained Self Supervised Depth estimation PackNet-SfM on Indian driving dataset and on Carla simulator dataset.

Graduate Research Assistant

Robotics Research Center, IIITH

Aug 2018 – Jul 2021 Hyderabad

Worked on the intersection of SLAM, Computer Vision, Deep Learning, and Robotics. Developed robust pose graph constraints using scene semantics and developed rotation invariant deep feature descriptors for feature matching.
Published in ICRA and IROS conferences.