In October 2019 the Web Dev club held a 3 hour Hackathon at the University of Victoria. This event was not a MLH event, it was independently supported and held. I entered as a solo entry and won 2nd place.
The goal of my 3 hours was to design the rough idea for a product and get a basic back end working. I was able to achieve both of these major goals.
Firstly, the product idea, finding seasonal movies is hard because you only think about them on a yearly basis. More often than not, consumers of movies know the actors they don’t like over the ones they like. Moreover, it is less likely to know what exact movies contain these actors until it’s multiple minutes. The idea is to pick out the actors you don’t like by their face and the software will omit the movies containing all actors entered.
The software executed perfectly in an alpha state. By using open source facial recognition libraries the main tech-debt for this project was offset into product development. As a primitive back end, the concept works. By placing images of actors faces into the root directory it will match the movie with a picture of them in the frame. Within the time constraints the technical overhead of this project was enough to get into the top 10 of the contestants.
In the demo period, the execution of the demo worked flawlessly. Though the first place prize was well earned, a goosebumps style choose your own adventure with text-to-speech. In summary, it was a fun event where a lot was learned.
-Python Image Library
BY RYAN BARCLAY