Sea-It

Inspiration

• We wanted to create a citizen- science inspired application

What it does

• Using a photo provided by a user, the app uses a Machine Learning algorithm to identify an animal in the photo. Once something is identified, the program then returns a confidence score to the user.

How we built it

• We used Android Studios to develop the application and Java and Javascript to write it. Using a machine learning algorithm from an open-source library, we used a trained model to classify ocean images and we integrated it with an Android mobile app. Finally, we used Firebase for the authentication.

Challenges we ran into

1. Initially, we started to try machine learning with Google Cloud, but it would have taken hours to train the machine, so we had to use a different approach.
2. Firebase was difficult to implement for the Google authentication.

Accomplishments that we're proud of

None of us had used machine learning before but we were able to integrate an image classification model with our app that works well. We are also very proud of how sophisticated our app ended up looking/feeling.

What we learned

• Backend development is very hard if you have only have little experience with it.
• Front-end development using XML was pretty intimidating when using it for the first time, but easy to implement once you get the hang of it.

What's next for Sea-It App

1. Continue integrating Firebase!
2. Upload to Google Play Store

Built With

• android-studio
• firebase
• java
• xml