Learn more about me
I am currently a Final Year Mechanical Engineering undergraduate [B.Tech] student at Indian Institute of Technology (IIT), Dharwad.
At the start of my undergrads, I was very much interested in the field of Robotics, so my major obviously became Mechanical Engineering. But soon, I realized that not only is Mechanical Engineering is important, as an interdisciplinary topic, even some knowledge of Computer Science stuff such as ML, DL, and Computer Vision is essential.
Projects
Lines Of Code
Hours Of Learning
Recommender Systems; Time Series Modelling; Natural Language Processing; Computer Vision.
Computational Fluid Dynamics; Turbulence Modelling; Finite Element Analysis; Numerical Analysis; Iterative Procedures.
Simultaneous Localization and Mapping; Forward and Inverse Kinematics/Dynamics; Multi Body Analysis; Path Planning.
Check Out My Resume
An aspiring young student with skills in Machine Learning, Data Science, Robotics and Simulations. I love to fiddle around with different ideas and I find that such explorations are a great boost for learning.Interested job profiles:
Awarded with scholarship by Govt. of India for top performance (Top 1000 in country) among all class X students in India
Joint Entrance Test, common for all the engineering aspirants in India. Achieved a top All India Rank (99.9%ile from 13 million aspirants for JEE Mains) to be admitted to an IIT.
Certified to be a 99th %ile IQ in the world. Scores indicate 153 on Stanford Binet IQ test with 15SD.
Indian Institute of Technology (IIT), Dharwad
Late P B Jog Jr. College of Science and Commerce
With Computer Science as specialization (In place of two other subjects like geography, biology etc.)
Jnana Prabodhini Prashala
Won school scholarship for speed reading as well as mental ability. Won Silver medal for Homi Bhabha Jr. Scientist Exam for excellence in scientific abilities.
PARI Robotics, India
Course Projects
TETRIS - Programmed the famous console game of TETRIS in C++ using OpenGL library. Added motion effects and 3D look to the Tetriminos (The pieces that fall down).
Recommender System for Movies - Implemented various ML models based on various methods of Content based Filtering, Collaborative Filtering, Matrix Factorization as well as Clustering techniques upon the MovieLens 1M dataset.
Analysis of MRI images - Analysis of MRI images in MATLAB as well as figuring out the critical regions of a scan. Performed analysis to determine the SNR ratios and the error rates.
Mechanical Analysis of Engine - Performed static and dynamic structural analysis of 4 cylinder IC engine to determine the effects of the cyclic motion generated fatigue on the crankshaft and the piston.
Neural Style Transfer on Text - Did I write it or Shakespeare. Implemented a NLP Deep Learning model that recreates the written text to stylize it as Archaic English (Or something else, depending on the training data).
Dynamic Modelling of a Vehicle - Programmed a vehicle in CARLA Simulator using python and simulated its movement, steering, traction and forces experienced by the vehicle in real life to simulate it's control for Autonomous Driving.
Other Projects
One of my most challenging projects, done in 1st year of undergrads. Built a robot that starts by random movement and over time learns the contruction of its body, developing a sense of motion and ultimately, walking in an ergonomic manner(gait).
Based on the research of Prof. Hod Lipson and Prof. Josh Bongard, one of the early influences that gravitated me towards the field of robotics.
Designed and created a Quadcopter alongwith my roommates. Made completely from scratch with a self programmed flight controller and modelled starting with flight dynamics and navigation system to fly unobstructed across the campus with a camera mounted to detect the hinderences in the path.
An extension to the BTP, extended the project by adding an AI brain to the car by implementing Segmentation CNN, U-net, to identify the road, the obstacles, pedestrians, and other vehicles. YOLO for Traffic Sign detection and identification and a Reinforcement Learning Framework for decision making based on the rsults of the environment perception to execute the necessary motion through the dynamic model.
Bult a web app using Flask framework in Python. Implemented web scraping with bs4 to create the training data, trained basic ML models in scikit-learn to predict the closing price of the stock for the next day.
Applied CycleGAN to convert horses into zebras and vice versa. Also implemented Pix2Pix alongside U-net Generator for creating sceneries in different sesaons.
Used various images of human faces to train a Deep Residual Neural Network to identify the facial features that denote the emotions.
Used a Convolutional AutoEncoder to train encoder-decoder pair that takes a low-res image as input and outputs a high-res image as output.
Implemented a pretrained GPT-2 model for transfer learning on various scientific papers to generate summaries of those using the training set generated via various sources on Github and Medium.
Implemented Tokenization on the Tweets data using count vectorization and applied various classification techniques such as Logistic Regression, Naive Bayes Classifier and SVMs.
Analyzed the characteristics of Flwmeter life cycle and identified the flowmeters near the end of lifecycle. Also, implemented a Timeseries prediction model for Life cycle analysis of an electromagnetic Pump. Data used was obtained from this and this datasets from UCI archives.
Contact Me
Pune, Maharshtra, India