1. Researcher Assistant - AI4CE Lab, NYU Tandon

   Working under the guidance of Prof. Chen Feng

   Jan'23 — ongoing
   Long Baseline Spatial Reasoning
   • Troubleshooted point cloud misalignment challenges by conducting meticulous analysis of ground truth data,debugging errors within the dataset generation pipeline
   • Implemented inverse perspective projection using corrected ground truths, leading to the achievement of accurate 3D visualization of indoor environments by automatically aligning individual point clouds
   • Migrated the existing point cloud generation pipeline from open3d’s conventions to habitat-sim conventions
   • Improved the accuracy of the existing dataset generation pipeline by calculating the Intersection-over-Union (IoU) between two generated point clouds
   
   Long-term Visual Place Recognition (VPR)
   • Utilized STEGO, an unsupervised semantic segmentation algorithm, to generate semantic data for NYU CARLA dataset
   • Optimized the data generation process using pytorch and openCV after rendering images using STEGO
   • Developed a python3 script to convert rendered images into videos by stacking all images into frames
   
   

   • Software Engineer - Deep Learning and Computer Vision, Siemens India

     Jul'21 — Aug'22
     • Engineered a computer vision system to implement template matching for the identification of electrical elements in Single Line Drawings (SLD)
     • Utilized object detection with the yoloV4 model to recognize electrical elements within SLD, achieving an impressive accuracy range of 85 - 96%
     • Designed and implemented an Optical Character Recognition (OCR) system employing pytesseract to discern the electrical attributes of elements by extracting information from object detection results. Demonstrated a commendable average accuracy of 90%
     • Established a Line segment detector pipeline for the detection of various breakers within a panelboard in SLD, leveraging tools such as Python3, and OpenCV
     • Integrated the object detection model, OCR system, Line Segment detector pipeline to intelligently establish correlations between various breakers within a panelboard and their corresponding electrical attributes
     • Implemented an AutoML system, to detect user-selected symbols in SLD as microservices, utilizing Flask and python, deployed on a Linux environment

1. New York University (NYU), Tandon School of Engineering, Brooklyn, NY

   Sept 2022 - May 2024

   Masters of Science - Mechatronics and Robotics Engineering
   GPA: 3.778 out of 4.0

   • Secured NYU Merit scholarship of $9000 for entire duration of study
   • Working as a student ambassador at NYU Phonathon, raising funds for the NYU community by reaching out to alumni
   • Volunteering as a tech specialist at NYU 3D print club. Hosted and organized workshops for undergrads on cad modeling using autodesk Fusion 360
   • In Fall 2022, I volunteered and guided PolyPrep Robotics Club's 20 students through the First Robotics Challenge, propelling them to the State qualifiers.
   • Coursework: Robot Perception (ongoing), Leadership in Organization (NYU Stern) (ongoing), Mechatronics, Advanced Mechatronics, foundations of Robotics, Robot Localization and Navigation
   • Technical Skills: Python, C++, Pytorch, Pandas, OpenCV, Open3D, Numpy, Docker, Scikit-Learn, MATLAB, Simulink,ROS(learning).
   • Non-Technical Skills: MS Office, Google Suite

2. Hindustan Institute of Technology and Science, Chennai, India

   2017 - 2021

   Bachelor of Technology - Mechatronics Engineering
   GPA: 9.09 out of 10.0

   • Gold medalist and second rank holder for the academic year 2017-2021
   • Participated and qualified for the finals of Smart India Hackathon - 2020. During this hackathon, I and my team worked on developing a CNN model for crop detection and recognition of 25 different crops using live camera feed
   • Represented HITS at Technoxian 2019 - International Robotics Championship . Participated in two events - Robo Soccer and, Fastest Line Follower. Qualified for the finals of robo soccer
   • Organized and Hosted several Machine Learning and Deep Learning workshops while being the Secretary of Artificial Intelligence (AI) Club
   • Coursework: Mobile Robots, Machine Vision, Embedded Systems, Microprocessor and Microcontrollers, Robotics and Machine Vision Systems, Artificial Intelligence for Mechatronic Systems, Rapid Prototyping
   • Technical Skills:Python, C++, Autodesk Fusion 360, Arduino, Tensorflow, 3D printing, Laser Cutting

Here's a collection of resources that played a crucial role in my journey of becoming a perception engineer. I hope these materials prove beneficial in your own learning adventure.

1. Blogs

   • YOLO V8 - A Deep Dive into its Advanced Functions and New Features
   • Depth Estimation: Basics and Intuition
   • Understanding Transformations in Computer Vision
   • Inverse Projection Transformation
   • Kalman Filter: An Introduction

2. Online Courses

   • Mobile Robots and Sensing by Prof.Cyril Stachniss
   • Robotics 501: Mathematics for Robotics by UMich
   • CS230: Deep Learning by Stanford University
   • Robotic Manipulation: Perception, Planning and Control by Prof.Russ Tedrake
   • CS549: Computer Vision by Prof.Nitin J Sanket
   • First Principles of Computer Vision by Prof.Shree K Nayar