Machine Learning
Automatic Traffic Light Detection System (ATLDS)
A common problem in self-driving systems; traffic light detection is a necessity in keeping drivers and passengers safe with autonomous vehicles on the road. Using tensorflow and keras in the Jupyter Library I was able to create two models that classified with over 99% accuracy with no misclassifications of red lights (the theoretically most dangerous mistake to make). This was all completed using only about 1500 sample images, and would be even more accurate with more data.
Locator:
This is a geotracking code that takes the user's IP address, which is then put into an API request at "ipapi.co", this returns the user's location including their latitude and longitude. The google maps integration takes that lat and long and instantiates a map centered on that location. This started as a way to visualize the amount and type of sensitive data you release about yourself online.
This code is also a great way to see if a VPN is working as advertised, just connect to a VPN and you should see your location change when you run the code again.
*2024 note: google styling wizard changed to paid model, expect basic map.
Skewmorphic Records:
created in p5.js, this audio visualizer code takes an audio file and outputs the amplitude over the duration of the song and applies it to an outward spiraling pattern. It also takes the album art (or other uploaded image) and makes the background the average color in the original image file.
Run it for yourself now: Click here
Bad idea right? : Olivia Rodrigo
Leave the Door Open: Silk Sonic
Kickstart my Heart : Motley Crue
Mr. Lonely: Midland
Octo: The Hungry Octopus Game
Using Pygame, created a game where the goal is to collect all items before time runs out. ATLAS 5650
