Currently looking for a full-time opportunity in robotics software: perception and sensor fusion systems in mobile robots and manipulation. I am also interested in working on real-world problems using computer vision algorithms for different fields of applications.
- Application of Moveit in ROS2 to understand skills and its use for extracting object information for manipulation
- Using VIO and behaviour trees for robot perception, navigation, and action planning in mobile manipulators
- Using BundleSDF for object pose extraction for gripping industrial parts with irregular object geometry.
- Experimenting with vision pipelines for use in a Voxposer setup
- Completed my Master's Thesis at Fraunhofer IPA, Stuttgart on Combining Vision, Force Control and Reinforcement Learning for the task of gear assembly
- Developed a multimodal system in the perception of a robot manipulator in an assembly line.
- Created a data-efficient system capable of training the Reinforcement Learning model quickly with quick deployability
I am interested in using learning-based systems with scene understanding for robots like mobile manipulators in the field of inspection and disaster rescue. For this objective, I have been developing my skills in manipulators, mobile robots and robot perception through internships and personal projects.
Some of the topics I have worked on are:
- Multi-sensor fusion
- Visual Reinforcement Learning
- Robot perception for navigation and manipulation
- Optical Flow, Visual Odometry, Depth Estimation
- Synthetic Data Generation
- Object pose estimation
- Mechatronic systems
- Robot body and circuit design and prototyping
- I am a Master Student of Robotics Systems Engineering from RWTH Aachen University, Germany
- I completed my Bachelors Degree in Mechanical Engineering from Indian Institute of Technology, Kharagpur
- Languages: Python(Pytorch, Tensorflow, Stable Baselines, Gym, MuJoCo), C++, Embedded C
- Tools: NVIDIA Omniverse, ROS, Docker, WANDB, MATLAB, LaTeX, Blender