Get Your Perfect Matches

Inspiration

Many of our peers struggle with getting the perfect outfit of the day, and this is especially problematic when we are running late for our morning classes. Why don’t we build a software that would generate you a perfect outfit where you only need to make a single choice?

What it does

The user can input a path of images of all his/her clothes, and an image of a piece he wants to wear. The algorithm will output another piece of cloth that will match with the piece that he wants to wear according to similar-color principle of fashion.

How we built it

We would first analyze the colour of the input image, and then analyze the color of all the images in the path. We would employ the main theory of fashion in order to satisfy the general population’s sense of beauty, where it will output the user a piece of clothes of similar color.

Challenges we ran into

We are all U1 students this year. We have limited programming experience and are new to AI as well. We’ve faced many problems in our code including importing the libraries, the unexpected errors in our program, and the challenge we face to create an algorithm that can detect the images’ color.

Accomplishments that we're proud of

The program properly outputs fashionable items that reduce the number of choices users have to make each day. It helps them to save time while dress properly for the day. Besides, this project is eco-friendly since by properly matching of clothes that users own right now, they get new lookings without the need of purchasing new clothes as often as they would do without this software.

What we learned

Sometimes debugging takes longer than we thought! We just need to be patient and believe that we can do it. Analyzing colors teaches us the importance of proper choice in statistics. We also gain a lot of experience in group work and collaboration.

What's next

The amount of AI-related ideas used in this project is a bit limited due to our plainly experiences of it. We would like to improve it by employing more AI-related techniques in the future.