SeeOn

Mobile app where users receive a personalized seasonal color analysis, clothing recommendations aligning with the palette of their color season, and can virtually try on clothing.

Comment

Inspiration

Seasonal color analysis has become a viral sensation across the internet. While there are several variations, it commonly categorizes a person based on their facial features into one of four seasons (Summer, Spring, Autumn, or Winter), with a further sub-classification of Light, True, or Bright. Knowing your season helps guide you to make style choices that are in harmony with your natural features and palette. However, getting a professional style analysis done in-person can cost hundreds of dollars. As somebody who loves exploring my personal style, I created SeeOn in hopes of providing a more accessible way for consumers to explore their color season virtually.

What it does

The user is prompted to upload an image of themselves, which is analyzed to assign them to one of the 12 color seasons. Their season determines their color palette, which is made up of main and neutral colors. They can then search for clothing recommendations online that align with their color palette. The clothing items can then be saved to their closet. These saved items can be tried on directly from the closet, generating an image of the user wearing the clothing item so they can see how it would look in real life.

How we built it

SeeOn leverages the YouCam AI Facial Color Tones Analyzer API to find the color of one's hair, eyes, and skin tone based off a picture. It then calculates the overall warmth, saturation, contrast, and brightness of these colors, which is then used to categorize the user into their color season. The clothing recommendation system uses the Channel3 API to search its database and provide the user with clothing options that match colors from their color palette. To allow the user to see how the options look on them in-person, the YouCam AI Clothes Virtual Try-On API is used to allow users to generate a realistic image of themselves wearing their chosen clothes. The mobile app was built using React Native.

What's next for SeeOn

I would like to continue improving the functionality and accuracy of SeeOn's features, especially the clothing search function. Currently, clothing search is based on the written description of an item's colors, not exact color values. This can lead to inconsistencies, where an item does not match the actual color the user is searching for (for example, a dark red and a bright red might both be considered "crimson"). Since the subtle differences in color shades is very important to seasonal analysis, user experience would be greatly improved if SeeOn could filter clothing items to make sure items displayed are truly of a similar color value. I think it would also be helpful to have an expert review, to improve the generated color palettes and help tweak the season analysis algorithm to ensure that the results are accurate. Finally, I would also like to build a more robust backend for storage of user data. Additional features I would like to explore include social features such as sharing your curated closet with friends, being able to import your own colors, and better curation and personalization, such as being able to search for specific silhouettes or aesthetics of clothing.

Share this project:

Updates