cv
This is My updated Resume
General Information
| Full Name | Abhinav Kumar |
| Location | Boston, MA |
| Phone | 857-693-9631 |
| kumar.abhina@northeastern.edu | |
| https://www.linkedin.com/in/abhinav-kumar-777b18211/ | |
| GitHub | https://github.com/kumar-abhina |
Education
-
Dec 2024 Master of Science in Robotics
Northeastern University (NEU), Boston, MA - Relevant Coursework: Reinforcement Learning, Deep Learning and Pattern Recognition, Mobile Robotics, Computer Vision, Numerical Optimization, Autonomous Field Robotics, Control Systems
- GPA: 3.8
-
May 2022 Bachelor of Technology in Robotics Engineering
SRM Institute of Science and Technology, Chennai, India - Relevant Coursework: Robotics 101, Machine Vision, Planning and Decision Making in Robotics, AI for Robotics and Vision
- GPA: 9.14
Experience
-
Oct 2022 - Current Research Assistant
Helping Hands Laboratory, Northeastern University, Boston, MA - Developed a few-shot imitation learning algorithm using Vision Transformers and Deep Learning, enabling robots to learn object manipulation from minimal demonstrations.
- Fine-tuned real-time segmentation and DETR object detection models in Detectron, achieving a 23% higher AP-75 score compared to Mask R-CNN.
- Implemented planar grasp detection using SE(2)-equivariant CNNs to improve grasp learning efficiency and real-time performance.
- Published paper: One Shot Imitation Learning via Interaction Warping (CoRL 2023, Arxiv: https://arxiv.org/abs/2306.12392).
-
Jan 2024 - July 2024 Robotics/Automation Engineer
NUVERA Fuel Cells, Boston, MA - Designed and developed an automated stacking system to categorize Off-Stack Fuel Cells (OFFs), increasing sorting efficiency by 40%.
- Programmed a UR10 robot-based pick-and-place system in ROS2 and MoveIt2, integrating a Cognex smart camera with OpenCV to boost accuracy by 35% and reduce misplacements by 25%.
- Developed Python scripts on Linux to automate system monitoring, logging, and debugging, improving deployment efficiency on NVIDIA Jetson hardware.
-
June 2021 - Oct 2021 Automation and Robotics Intern
FESTO India, Chennai, India - Optimized industrial electro-pneumatic frameworks, reducing equipment cycle time from 12 seconds to 8 seconds per operation.
- Simulated diverse industrial manipulator operations using CIROS; designed control parameters that reduced production time by 11%.
-
July 2020 - Dec 2020 Robotics Software Engineer
DRDO Lab, Hyderabad, India - Developed a 4WD robot for lane tracking and autonomous path following via ROS-based object detection and SLAM, cutting navigation errors by 50%.
- Integrated IMU and GPS data with an Extended Kalman Filter (EKF), achieving a 15 cm RMSE in drone position estimation.
- Created a real-time autonomous quadcopter landing system utilizing fiducial markers and pose estimation, improving landing accuracy by 25%.
Academic Projects
-
Sept 2024 - Dec 2024 Structure from Motion using Bundle Adjustment
- Implemented classical 3D reconstruction on an NVIDIA Jetson Nano with ~100 images.
- Optimized re-projection error from 600 px to 3 px using Cholesky-based BA.
-
Sept 2023 - Dec 2023 Image Mosaicing and Depth Estimation
- Formulated a panorama-generation algorithm for 25+ images using feature detection, RANSAC, and Homography.
- Conducted multi-view stereo depth estimation with monocular cameras and pose-graph optimization.
-
Aug 2022 - Dec 2022 Autonomous Navigation and SLAM
- Developed an autonomous robot for reconnaissance in simulated disaster zones, employing Cartographer SLAM and frontier exploration.
- Achieved 87% success rate in victim detection using extrinsic camera calibration.
Skills
-
Programming Languages
-
Software and Frameworks
-
Libraries and Tools
-
DevOps & Cloud