Fashion isn't easy. Yet it's so critical to first impressions. Diverse dress codes and rapidly changing trends make it difficult to figure out what to wear. Stylezure is your personal stylist, helping you look your best, anytime and anywhere.
What it does
Stylezure is a chatbot. Given the event you're attending, and a picture of the outfit, it suggests changes to your outfit.
Stylezure is a chatbot. Talk to him, and he figures out what events you're going to and helps you pick a outfit for it. Stylezure then looks (literally) at you, and gives you instant feedback on your outfit. Stylezure then suggests potential fixes, as well as
How we built it
The chatbot was built using Microsoft Azure Bot Service API. Queries were processed using Microsoft's text analyzer. Pictures of outfits were analyzed using IBM Watson Visual Recognition and Google Vision. Suggestions for clothes were curated using Amazon.
Challenges we ran into
Providing suggestions for a better outfit based only on a picture proved challenging. Determining which outfits would be best for which situations was also a challenge.
It was difficult to use images with some of the APIs. We resolved that by standarizing the image into a flexible format usable by all the various APIs.
Accomplishments that we're proud of
Building a complete chatbot that effectively solves a common problem. Being able to split a large project into many parts, while maintaining cohesive teamwork.
What we learned
We learned how to leverage many APIs to work smoothly together, rather than collide with each other. We learned how to train chatbots to respond intelligently and naturally to human input.
What's next for Stylezure
Stylezure will train itself on many more features/aspects of fashion, like color and rarer/hipper styles. Furthermore, Stylezure will recommend clothes using other useful statistics, like wardrobe content, latest store discounts, etc.