NOTE: Our website would have been (haha get it, cause auta is the project name), but we ran into issues that the staff could not resolve. So we are using github hosting, and you can find the link on the bottom.

All four of us are studious high school students, and we all struggle with focusing while doing homework. As this is a common issue among all students, we wanted to make an app to combat this issue.

Our project includes a chatbot and a focus detector. The chatbot is a place where you can express your feelings and talk about your day. Our chatbot will automatically respond with a message based on the mood of your messages, and then suggest something to brighten up your day if you are down. Our focus detector is something the students can use while doing homework, to enforce focus while doing homework and help build great focus habits.

We built the front end with react-native, and then used golang and java as our backend. For the chatbot, we created an http server that returns the emotions of any given sentence using sentiment analysis. We also created Java modules for react-native to handle the machine learning for the focus detector. We used a CNN made in Tensorflow and trained it 10k+ images of team members' faces.

One challenge we ran into was we couldn't add the react-native bindings for openCV because the process was too lengthy and added too many dependencies.

We are proud of our stunning cross-platform user interface, and our machine learning backend for both the chatbot and the focus detector.

What we learned: creating disposable HTTP servers, training neural networks, creating apps in react-native

Whats next: -Implementing openCV, which will result in better detection of focus

Share this project:


posted an update

If you guys are wondering why it says 30,000+ lines of code, we rechecked, and it is actually just around 1000+ lines of code (excluding outputted lines from the jupyter notebook). We told this to judges, so don’t worry

Log in or sign up for Devpost to join the conversation.