The problem is many of us have a short fuse when it comes to shopping, especially men. You don’t want to waste any time searching trough all the options, and you might not know what looks good with what. There are always occasions where you don’t know what to wear or how to complete your look. We’ve worked on solving these common problems among men between the ages of 21 and 38.
We’ve found ways to accommodate your style trough the Zleek app. We inspire, educate and accommodate you through styles and choices that will fit the type of person you are. For Zalando, that means more sales, and for the consumer it means an effortless shopping experience. The secret to impeccable style is knowing how to coordinate the colour of your clothes and dress for your body type. Our app is the middleman between you, the customer, and the online shop. The app acts as the most helpful shopping advisor as it demonstrates options that fit your signature style, based on the picture you take and your own input.
What it does
We’ve found ways to accommodate your style trough the Zleek app. The app acts as the styling advisor specifically to help match clothing items to curate a look by taking the persons style in mind and find clothing for your body shape.
How it works
You will have to pick a style out of four general styles that you feel related to. Take a picture of your outfit and highlight a specific item you want to either replace or want to match it with other items to buy a look. .
Based on colors and shape of the clothes you are wearing, we can present the ideal items that will suit your current wardrobe as well as interest you in something stylish you didn’t even know you would love.
With the easy touch of a button, you will be able to proceed to check out and safely transact your order.
How we built it
We focused on the difficulties in the project and started solving them at first. For example for extracting data from an image in terms of garment, color and pattern. We didn't succeed in implementing a very useful clothing parsing library, we found this research paper that has an implementation of the solution. With some more time and an Unix based development environment we are sure this is implementable in a short notice.
Suggesting garments to a user was the most difficult part for us. It was important for us to know the fundamentals of fashion consulting and put those rules into code. During our research in this field we came across various set of rules that we can use. We focused on color combination styling and wrote the set of rules that lead to a more stylish look within your own style category. Other possibilities that we didn't implement in the designs but have designed in terms of code is clothes suggestion based on your body build.
Challenges we ran into
We identified 2 main challenges.
- Classifying garments through an image and finding colors and patterns
- Making suggestions based on what the person is wearing.
Accomplishments that we're proud of
We proudly converted the basic fundamentals of how a fashion consultant gives advice. By mainly focusing on color matching. From there we are able to build further with body shape.
Even though we didn't have something working that extracts data from an image, (data like colors and patterns in clothes). We got closer than people expected we could come.
What we learned
Leo: java, playframework, matlab, opencv Manish: Fashion consulting, matlab
What's next for Zleek
- Narrow down clothing options for customers while having better clothing suggestions. We can give more accurate suggestions if we have time to do proper research on fashion consulting.
- make better suggestions by analyzing previous preferences and purchases
- ability for user to share the suggestions made by Zleek and run a poll to get to a perfect find.