Ever faced the problem of not knowing what to wear for events like interviews, career fairs, or just on a daily basis? That's where our application comes into play -- having a virtual closet that can provide an outfit in an instant would be able to give anyone a good starting place in deciding what to wear -- and you wouldn't even need to move a finger!

What it does

The app can identify whether the user wants a casual outfit or a formal outfit and then generate a combination of clothing to make an outfit. Each time the user says something with keywords "casual" or "formal", the app will generate a new outfit for the user to wear based on whatever is in the current database.

How we built it

Using a Dialogflow agent and a Firebase realtime database, we were able to write and read articles of clothing from a virtual closet database using the Google Assistant as our interface. Using javascript, we implemented our own backend to process items in the database and generate an outfit when the user asked for one.

Challenges we ran into

We had a lot of problems dealing with the Firebase realtime database because it was different what we were used to and we didn't have much experience in javascript or databases, so it was new to all of us. Because of that, we were also unable to implement user input being added into a database.

Accomplishments that we're proud of

We were able to figure out the Dialogflow agent after a lot trying and retrying which was extremely satisfying, especially because we all little experience using Dialogflow and Firebase. We also had to familiarize ourselves with javascript since it was the language that was used by Dialogflow.

What we learned

We probably aimed too high and too broad at the beginning for our original idea and we found that we ran into many bumps along the way. By the time we scoped down our problem, we also had too little time left because we had spent too much looking at too many things that we ended up not implementing (like different web APIs). We also should have spent more time getting familiar with Dialogflow and Firebase earlier on because our familiarity and ability to use these interfaces were probably the biggest challenges that we faced.

What's next for Wardrobe Assistant

What we were hoping to implement in our app was the ability to take in user inputs. Currently there is only a preset database that the app draws articles of clothing from to generate outfits. If users can input their data into their own personal database, our app will become much more personalized and also much more useful in each person's daily life. We also want to improve the quality of our data; currently we are using a Firebase realtime database that holds all the information but ideally, we want user input to be added into one complete database and then have another database hold the keys for casual or formal items, making it even faster to access each piece of data. We want the user feedback to be used as data points to help make better suggestions for the user in the future as well. Ideally we would also implement a more in depth and detailed database with more tags that also includes images for the user to make it more user friendly and also implement a weather API to help make more pointed outfits depending on season and temperature.

Built With

Share this project: