Inspiration
Have you ever been on a walk and found a super cute dog and wanted to know the breed of it and where you can adopt a similar dog?
What it does
It's a web app that allows a user input an image (jpeg) of a dog they saw. It uses a machine learning model to identify the breed of the dog, and then uses that information to request nearby dogs looking for homes using the PetFinder API
How we built it
We built a backend in Python using Flask and a frontend in javascript/html. The machine learning model is based on the Google Inception Model and trained a few head layers upon this model using the Stanford Dog Breeds dataset to differentiate between dog breeds in images. This information is passed to the PetFinder API to find the nearest dog of that breed looking for a home.
Challenges we ran into
We first started developing an android app, that none of us had experience in. Trading that in for a web app made it significantly easier and was more in our wheel house. After that it was using flask as our micro-network server which came with some really confusing documentation.
Accomplishments that we're proud of
Finishing the project! Using machine learning and making a nice web page.
What we learned
HTML/Python piping and Python/HTML rendering
What's next for Dog-go
Include other APIs to expand capabilities, like Amazon or Google Maps.
Log in or sign up for Devpost to join the conversation.