cv

PhD Mathematics

• Computer Vision
  • Advised by Dr. Prashant Athavale

MS Mathematics

• Research in Variational Methods for Image Analysis

BS Statistics and Mathematics

• Computational Mathematics

Assistant Professor in Data Science and Statistics

• Developed and instructed a course in Data Visualization with Shiny
• Taught Base SAS and python for data analytics and statistical methods
• Taught Differential Equations and Linear Algebra for engineers
• Supervised students in Data Science projects

Statistician (Intern)

• Studied sampling plans for assessing and quantifying lot quality
• Cleaned and compiled change control data from all departments for further analysis
• Performed survival analysis on change control data and suggested improvements to manage and streamline the process
• Learnt how to perform experimental designs (DOE) for bioassays in JMP

Data Scientist

• Performed data cleaning on historical student data to make it ready for analysis
• Analyzed student retention & graduation data in R to identify significant factors that affect retention
• Built predictive models to identify high risk students
• Presented findings to university administration, for use in student risk monitoring

Machine Learning Engineer

• Mined and analyzed historical weather & sales data, performing feature selection for model training
• Cut waste by building predictive models for foot traffic to inform staff scheduling and food production

Managing Instructor

• Created an optimized scheduling chart to increase the service capacity
• Managed the instructors, and oversaw day-to-day operations

LoRA as a Flexible Framework for Securing Large Vision Systems

• In this project, we apply low rank adaptations (LoRA) for building security patches for vision attacks in large vision models
• We show that we can compose multiple LoRA patches to address multiple vision attacks without the need to retrain the entire model

Image Processing Application

• In this project, I make use of the flask framework in python to build a web application for processing images
• The app harnesses the opencv module for editing. Including;
  • Face Morphing
  • Image Restoration (Denoising & Inpainting - Object/Background Removal)

PyEBSD-SEM

• Created a python library for accessing, measuring, visualizing and restoring orientation maps from scanning electron microscopes.
• Used deterministic models, ML models, and hybrid models as my main PhD contribution.
• Created python access scripts for Dream3D software piplines.
• Proposed 3D convolution kernel to estimate texture with spatial deviation in grain microstructures

Evaluating Intervention methods in STEM Courses

• This project was to compare student by demographics and test scores to determine whether the interventions implemented by the college in previous years were effective in increasing student perfomance (measured using graduation rates)
• Next, our goal was to determine which factors are fundamental in predicting the success rate of a student, so that the college can put in place measures to help weaker (students at higher risk of not graduating)

Better Scientific Software Fellowship

• The BSSw Fellowship program gives recognition and funding to leaders and advocates (Fellows and Honorable Mentions) of high-quality scientific software

SIAM MS24 Travel Grant

• Award to attend the conference and give a talk
• Emmanuel Atindama
• Role: Primary Recipient
• Amount: 600 (USD)

NIST Grant (2021-NIST-MSE-01)

• Award to develop Restoration of Orientation Data from Polycrystalline Materials Methods
• Prashant Athavale, Emmanuel Atindama
• Role: Graduate Student Researcher
• Amount: 30000 (USD)

Reconstruction of Grain Orientation Data Using UNets

Crystallographic Data Restoration Using Weighted Total Variation Flow and Hybrid Deep Learning

Measuring Impact of Student Success Retention Initiatives for Engineering Students at a Private Research University

Weighted Total Variation flow for Crystallographic Data Reconstruction

Weighted Total Variation flow for Crystallographic Data Reconstruction

Weighted Total Variation Based Algorithm for Reconstruction of Grain Orientation Data [A Comparative Study]

Partial Differential Equation Based Methods for Reconstruction of Grain Orientation Data