FashIntell

FashIntell is a platform for personalized outfit recommendations based on the user's preferences. Using a neural network, each user receives a carousel of outfits to dislike, like, or save.

Comment 1

Inspiration

Our main inspiration with FashIntell was our own experiences with online fashion. Although there are many apps that tailor a feed for you, like Instagram or Tiktok, we wanted to create one that was more focused on personalizing fashion for the user.

What it does

FashionTell is a website that displays image cards of outfits. Each card allows the user to dislike, like, or save an outfit. Each time this is done, a neural network is run on the choice, and the next recommendation is chosen for the user.

How we built it

We built the webapp using the React JS library. The main frontend was made using JavaScript, HTML, and CSS. This frontend was connected to the backend using Firebase, Google's cloud storage service. The backend consisted of the neural network, made with TensorFlow using a Google Colab file trained on this data set: link, and the data stored in the Firebase Firestore.

Challenges we ran into

Throughout the project, we ran into a number of issues. On the data side, we were unable to get the dataset we wanted since it required an approval, which still hasn't come through. On the technology side, we ran into a number of issues with Firebase. Firebase has recently been updated, making much of the old syntax obsolete. This meant that we were on our own to make it work. Between cloud functions and Firebase Authentication, we definitely had our work cut out for us.

Accomplishments that we're proud of

While our project is still unfinished, we're proud of how it turned out. We put in a lot of effort to make the neural network work and we are particularly proud of our frontend.

What we learned

We definitely learned a lot during HackIllinois. For one, we learned a lot about technologies we did not know before (Many of our members had never used Firebase, React, or TensorFlow before). Working in a team was also a learning experience. We had to learn how to use our talents in the most efficient manner to finish the project on time, and we also learned a lot about each other throughout the Hackathon.

What's next for FashIntell

Beyond HackIllinois, we hope to finish the webapp for a more commercial setting. We feel as though the webapp is marketable towards businesses that want their clothing items to be recommended or advertised to users, and hope to capitalize on this fact as FashIntell becomes a fully-deployed website.

Built With

Share this project:

Updates