This project is to create an as-yet unnamed tool to assist transgender people in voice training. Voices have many factors that affect their perceived gender. Some, such as pitch, can be measured easily, and there are tools to measure and track pitch changes. Others, like resonance and vocal fry, are difficult or impossible to measure directly. I am creating an app to track all of these factors, using machine learning to model the characteristics that can't be easily measured. The app is still in development and not yet ready for beta testing.
- • Discovered and researched a market opening for a better voice training tool focused on transgender people
- • Developed a Convolutional Neural Network machine learning model with Python, Numpy, Pandas, and Tensorflow to model vocal pitch, resonance, and several other gendered vocal characteristics better than competing apps
- • Developing an application in Javascript and React Native to publish
NHAT was developed by Olin College Students (including myself) as a Senior Capstone Project in 2023. National High Injury Network Analysis Tool (NHAT) is a free national tool that allows communities to generate interactive HIN maps and understand the relationship between safety, equity, and other contextual factors.
- • Researched user needs and designed an application to fulfill them
- • Developed an algorithm using OSMnx (OpenStreetMap data), NumPy, Pandas, SciPy, and Shapely to find the most dangerous roads in an area
- • Worked in a small team using an Agile development process
- • Developed a full-stack web application using Flask interfacing with an AWS database
View Website
View Repository
For our Astronomy and Statistics class, we had a project to do some statistical research given a dataset of lunar craters and write a scientific paper about it. I chose to research crater density at different latitudes of the lunar surface.
View Manuscript
For our Principles of Engineering class, we had a project to build a robot. We decided to create a robot that draws without using a gantry system. I worked on programming a website to convert an image to instructions, and using those instructions to run the robot.
View Website
In Summer 2020, I researched usage and equity of shared mobility systems in Providence, RI. We used R to analyze data from the systems, and created a Shiny app to allow Providence city planners to explore the data.
View Repository
For Data Science, our group decided to analyze racial disparities in COVID cases. We found a dataset with racial data on COVID cases, tests, and deaths, and analyzed its trends and uncertainly. We concluded that there are racial disparities in COVID data, but incomplete data could be obscuring more information.
View Report
More on GitHub