Shaelynn Brown

Computer Science student from University of Victoria

Portfolio

viewing featured..

Projects

ASL Recognition using Deep Learning Dec. 5, 2018

Python Pytorch

For my Honours Project, I am building a network to classify between 50 different sign glosses. The network is a 3D Residual Network. Currently, on my test set I am achieving 50% accuracy and I am looking on ways to improve it.

Chord-Flask A distributed hash table with a visualizer Nov. 24, 2018

Python Flask Typescript

Implementation of the Chord DHT Peer-to-Peer algorithm using Flask. For testing and educational purposes, there is also visualization of the network and key retrieval using D3

Day Trading System Scalability Project April 15, 2018

GoLang Redis Docker

This was a group course project for SENG 468 (Software Scalability Systems). The challenge was to develop a day trading system where user's buy and sell stocks. By the end of the semester, the system would handle 1000 concurrent users. My personal contributions to this project was setting up Docker Swarm deployment, configuring our load balancer, and writing logic for the transaction and quote server.

Dash Bot NwHacks 2018 Jan. 15, 2018

Python Flask

Dash is a brand new, next generation mobile banking app to help users navigate their banking experience more easily. It accomplishes this by offering users the ability to converse with it as it is a conversational bot. When logging in, users have the option to securely access their accounts via unique voice recognition. Specifically, I worked with the cognitive services API. I wrote the command server which handled the natural language processing using LUIS. I also wrote the voice authentication API in the auth server where you can create profiles, enroll profiles, and use identification against other user's voice profiles

Drum Detection Generating annotated training data Dec. 11, 2017

Python Flask SciKitLearn

This project is a software tool to generate ground truth drum event annotations from polyphonic audio. We aim to reduce the time it takes to gather ground truth drum event annotations. The tool will estimate drum events in polyphonic audio using a trained model and then allow for manual correction. The application can be used as a semi-automated tool for generating more data in order to improve its own model further and generate data for future research. By building a visual tool, we can visually analyze the accuracy of the predicted drum events, manually fix the errors, and use the new ground truth to retrain the model.

About Me

My name is Shaelynn Brown and I am currently finishing my last semester at the University Of Victoria. I will be graduating with Honours in Computer Science with a specialty in Software Engineering. Even though I love beautiful, friendly user interfaces, I get the most excited about using algorithms and data structures to solve problems. My favorite experiences from my undergrad include working awesome internships, competing in hackathons and coding competitions, and volunteering for Ladies Learning Code and Women in Computer Science and Engineering (WECS). Heading into the workforce, I aim to be a well-versed full stack developer, and hope to experience many different types of technology.

Contact Me