Portfolio
My Proudest Projects
Remote Control Water Turret
Originally created with the intention to combat my cat’s attempts to make me lose my security deposit by tearing up my carpet.
The turret was created to be accessible by any device on the LAN, this was accomplished by using a web interface that updated with the current video feed supplied by the onboard ESP-32 CAM module.
Anyone on my network could easily access the turret simply by entering the private IP address of the turret into their web browser. This allowed access for when I was working in my office on my computer, and also when I heard a disturbance when I was trying to sleep on my phone.
Control of the turret was then established by clicked on several buttons on the webpage, allowing the device to turn horizontally and tilt vertically. Finally there was a button labelled “Fire” that would activate the servo attached to the spray bottle, drenching my cat and discouraging his behavior.
RESTful API Receipt Processor
A more modern and technical backend programming challenge I imposed on myself was the creation of this Web API.
The Web API was created to simulate was many modern applications as possible, including the ability to call itself, handle complicated and malformed JSONs, and allow easy access for functionality.
Additionally the Web API was my first foray into creating large scale usable programs with GO, which quickly became my favorite programming language for any backend projects I would find myself working on in the future.
The program would take a “Receipt” JSON and deconstruct it into parts, creating an internal score based on certain arbitrary features (Date, Items purchased, Price of items, etc.) And return a unique value based on those variables.
Python Facial Recognition App
An App programmed in Python that allows for users to create and train their own facial recognition models using the Python facial recognition library and PyQT for the GUI.
The program was created with the intention of proving to myself that I could in fact use the PyQT5 library with Pillow, as well as facial recognition on top to keep it interesting.
The program turned out way better than I originally expected, it was fully capable of learning a persons face from just a few images, and was very easy to add new datasets using the GUI I created in QTDesigner.
One of my favorite uses of the program was placing images of Jerry Seinfeld in a crowd and seeing if the facial recognition library would be able to accurate discover and label him. It did so with flawless accuracy, though I only used high quality images and only ran a few tests.