Kulendu Kashyap Chakraborty

I am a Junior Project Attendent at the 3D Computer Vision Lab, IIT Jodhpur where I am advised by Prof. Avinash Sharma. Prior to this I completed my undergrad in Bachelor's of Technology (B.Tech) in Computer Science and Engineering in 2023.

I was fortunate to have been advised by fantastic researchers during my undergrad. I worked as an undergraduate researcher (intern) at the SWAN Group, Georgia State University,, advised by Dr. Ugur Kursuncu. Where I was engaged in studying human behaviour and intelligence on social media, integrating it with AI approaches to study and investigate their nature; I have also worked at NITTTR, Kolkata as a research intern .

My interest lies in broader areas in Computer Vision, Computer Graphics and Deep Learning. My current research focuses on studying various aspects of Computer Vision and Computer Graphics, strongly Digitizing Humans, Motion Capture, Deep Learning, and Dimensional Image Processing, especially with a focus in the space of 3D Vision - RGB-α, (x,y,z,θ,Φ), 3D reconstruction and Motion Transfer.

Email  -  Resume  -  Google Scholar  -  Linkedin  -  Twitter  -  Github

• October, 2023: Joined the Dept. of Computer Science & Engineering, IIT Jodhpur as a Junior Project Attendent
• June, 2023: Submitted and Presented undergrad final year project - Slides
• May, 2023: I will be attending 3D Vision Summer School, 2023.
• July, 2022: Thrilled to be selected as a participant at HackZurich, Europe's largest Hackathon.
• January, 2022: Excited to join the SWAN Lab (virtually!) at Georgia State University, as a Research Intern.
• Decemeber, 2021: Our paper accepted at Elsevier: Physiological and Molecular Plant Pathology.

My main research interests involves applied Computational Deep Learning & Computer Vision. I focus on exploring domains that involve excessive appilcations of problem solving with AI integrity in it, especially Deep Learning & Vision. I am generally interested in how to make train, train, and optimze better Computer Vision Models, out-performing in real-world parameters. Ocasionally I explore domains that are outside my major area of focu such as Linguistic Optmization in Natural Language Processing (NLP)/ Natural Language Understanding (NLU), and working as well studying the behaviour of LMs.

An AI based model is proposed for optical potato leaf image analysis for early and late blight disease identification. A benchmark dataset Plant Village is used to perform this study. Analysis is being performed to compare four deep learning model namely VGG16, VGG19, ResNet50 and MobileNet. The best model VGG16 is selected and tried to enhance its performance by parameter tweaking.

Given a multi-band 3D scan of the brain (preferably MRI scans), the aim is to semantically segment the Tumor from the volumized layers. Using 3D U-Net as the baseline architecture, all the pixels are semantically segmented to produce the Masks for the provided multi-channel MRI input. BraTS 2020 Challenge Data has been selected as the 'Benchmark Data' for analyzing & performing further operations.

Building algorithm architectures for detecting and analysing Collusiveness in the websites. Extracted and grouped the “Topics” from the HTML files using Latent Dirichlet Allocation (LDA) modelling, and also improving the overall accuracy score.

DrawIt is a screen on air that allows you to write/draw anything while just waving your fingers over the screen.

Movie-time is a movie review recommendation system that can say whether the review of the movie you wanted to see is positive or negative. It can basically help you to dicide whether the movie you wanted to watch is good enough to watch or not. Moreover it can save your time by just seeing the review, rather than the primitive method to watch the enitre movie and say is it good or bad.

Detecting flairs on Reddit. Flairs are basically like the tags that is used for filtering contents, resources etc. These 'flairs' is used in subreddits. .

Kitabe (Book in Hindi) is a Book Recommendation System built for all you Book Loversbook. Simply Rate star some books and get immediate recommendations tailored for you.

A Flask based web application to predict breast cancer.

Presented course content on the topics:

• April, 2023: CSE 1818PE52: Speech and Natural Langauge Processing - Morphology [Slides]
• November, 2022: CSE 1817OE21: Machine Learning - Decision Trees [Slides]

Miscellaneous stuffs → Photography | Sketches

[Web-Cite]