About

I am a Senior year student at the Department of Computer Science and Engineering, Jalpaiguri Government Engineering College. My primary research interest lies in Deep learning and Computer Vision. I consider myself as a lifelong learner and as one with longterm goals. I love to explore whether it is places or professional work. I love collaborating with diverse teams to solve intriguing challenges. I am passionate about Data Science, GPU’s (Of course) and Anime (lol!).
To know more about me, scroll down…

Interests

  • Machine Learning
  • Deep Learning
  • Image Processing
  • Android Development

Experience

  • Vir Softech Pvt. Ltd.

    June 2018 - July 2018

    Developed an application that extracts values against each fields of an Insurance form using Image Processing and OCR Engine

  • Tenpi Technologies Pvt. Ltd.

    May 2017 - July 2017

    Developed an Android App with features such as- user signup/login , Bluetooth Low Energy (BLE) connectivity and Location Services.

Projects

Predict Energy Efficiency of Buildings

Predicting the Energy Efficiency (Heating Load and Cooling Load) of buildings using Azure Machine Learning Studio.
Boosted Decision Tree Regression method is used as the model due to its ability to learn non-linearity in data.


Media Server

A local network based Media Server built on Raspberry Pi for seamless media streaming across devices (cross platform).


Py-Pen

An Open CV project that captures webcam feed and tracks a particular color and draws the path of that color movement. It can be used to write on screen by just moving a pen in air.


Knowledge-Based Expert system for diagnosis of Agricultural crops

The project uses a Convolutional Neural Network that is trained on a dataset of 40,001 images of diseased leaves belonging to 9 agricultural crops. This project deals with the development of an expert system to detect and identify the major diseases that some of the agricultural crops suffer from, and also suggest the steps to be taken after that. The model achieved an accuracy of 98.48% .


Contact