This is a spin-off company of my former employer DFKI. For past 3 years, I take care of computer vision, machine learning model implementation on edge devices (for robots) and the whole perception pipeline for agriculture robots. My current work includes laser weeding, plant growth monitoring and plant health monitoring.• Conceptualization, development and implementation of an instance segmentation approach for plant recognition• Continuous deployment of network architectures and… Show more This is a spin-off company of my former employer DFKI. For past 3 years, I take care of computer vision, machine learning model implementation on edge devices (for robots) and the whole perception pipeline for agriculture robots. My current work includes laser weeding, plant growth monitoring and plant health monitoring.• Conceptualization, development and implementation of an instance segmentation approach for plant recognition• Continuous deployment of network architectures and parameterizations using Pytorch, Tensorflow and Python• Created two datasets : 3D plant instance segmentation and Temporal growth stages for plant instance segmentation• Sensors selection, sensor position on the robot and evaluation of the sensors• Deployment of inference networks on the autonomous robot using ROS2• Retrieval of new recorded ROS bag files and annotation of the same• Coordination of the collection of large training data sets for different crops• Field planning for plant growth analysis Show less

* Development of plant instance segmentation machine learning pipeline for autonomous robots using Pytorch, Tensorflow and Python* Creation of 2D and 3D plant instance segmentation dataset without requiring manual annotations* Outcome of the work auto-labelled 10x more images than available annotated data* Development of a machine learning pipeline to assist manual labeling* Implementation of machine learning models on edge devices like robots* Successfully ran inference… Show more * Development of plant instance segmentation machine learning pipeline for autonomous robots using Pytorch, Tensorflow and Python* Creation of 2D and 3D plant instance segmentation dataset without requiring manual annotations* Outcome of the work auto-labelled 10x more images than available annotated data* Development of a machine learning pipeline to assist manual labeling* Implementation of machine learning models on edge devices like robots* Successfully ran inference models on agriculture robot Show less

* Researched towards explainable AI to analyze and visualize deep neural networks* Was developing a tool to help the machine learning community analyzing networks using different techniques such as CCA (Canonical Correlation Analysis), SVCCA (Singular Value CCA),PWCCA (Projection Weighted CCA) and CKA (Centered Kernel Alignment)* Researched focusing on the topic : Do different neural networks learn the same representations and to what extent they learn same things

* Tutored bachelor's and master's students for the subject - Introduction To Artificial Intelligence And Logic Programming* Designed the online tutoring structure as the COVID-19 happened right before the semester began

* Expertised in Mainframe which is used for large-capacity computers' programming used by mega-corporations for bulk data processing* Developed Mainframe programs from scratch and scripted multiple automation tools using REXX programming language, resulted in saving huge costs and manual efforts* Increased project revenue by 400%* Marked excellent production code deliveries involving million dollar calculations for the US retail and insurance domain clients