Style Transfer Using Diffusion

2023

Style Transfer Using Diffusion *We fine tuned stable diffusion to a Calvin and Hobbes dataset. We won the EE 371Q Digital Image Processing Ram’s Horn best project award!! EE 371Q is a famous class taught by Prof. Alan Bovik at University of Texas at Austin.

Technology used: Python, HuggingFace, OpenCV, Numpy*

Repository   |   Video   |   Report

LLMs Powered Egocentric Image Annotation Web Application

2023

LLMs powered Egocentric Image I developed a web application in Flask where a user can log in, and perform studies. The user gets an egocentric image from a camera wearer. Along with the images, the user also gets a list of possible activities that the camera wearer might be performing. The suggestions are powered by large language models (LLMs). The user can select the activity that best describes the image.

Technology used: Python, Flask, HTML, CSS, JavaScript, Bootstrap, MySQL, HuggingFace, Pandas, Numpy

Repository

Machine Learning to Detect Cravings and Stress in Patients With Substance Use Disorder

2019 - 2022

Wearable Sensor Framework

I worked with the team at UT Tyler and RAE Health(https://www.raehealth.com) to develop a machine learning algorithm to detect stress and cravings in patients with substance use disorder. We published in three major venues.

Technology used: Python, MATLAB, Pandas, Numpy, Scikit-Learn

Article 1   |   Article 2   |   Article 3   |  

EIT Toolkit

2021

EIT-Kit

I worked with researchers at MIT to benchmark different image reconstruction algorithms for electrical impedance tomography (EIT). I implemented image reconstruction algorithms in the toolkit called EIT-KIT. Our team submitted a strong rebuttal at UIST and published it.

Technology used: Python, Swift, XCode, Numpy, Pandas

Echolocation Robot

2020

Echolocation Robot

At UT Tyler, we (team of 3) built a robot that can avoid obstacles using ultrasonic sound emitter and receiver.

Technology used: Arduino UNO, C

Video   |   Repository

Machine Learning Algorithms for the Prediction of Bradycardia Risk in Preterm Infants

2019 - 2020

Infant Apnea

I worked on developing machine learning based risk stratification algorithm to predict life-threatening events in preterm infants.

Technology used: MATLAB, Python