Aubie or Not

Inspiration

I was inspired by my Deep Learning class I'm taking here at Auburn, and I wanted to put those skills to the test. I wanted to make an app that has something to do with Auburn. I combined those aspects to create an image recognizer that can tell if Aubie is in the photo.

What it does

This app analyzes a photo that you upload, and runs it through a machine learning model in order to tell you whether Auburn's mascot, Aubie, is in the photo.

How I built it

I built this app using Android studio, and I coded it in Java.

Challenges I ran into

The main challenge was figuring out how to create the model that could recognize Aubie in a photo. I had to find lots of pictures of each of the things I wanted the model to recognize. After that, it was a challenge to get the model to function in the app.

Accomplishments that I'm proud of

I'm proud that I was able to have a vision of a project and see it all the way through. Also it was fun to play around with Tensorflow to get it to recognize whatever I wanted.

What I learned

I learned how to train a machine learning model using Tensorflow, and then utilize that model in an android app for image recognition.

What's next for Aubie or Not

I could train the model to recognize every mascot from the SEC!

Built With

• android-studio
• java
• python
• tensorflow