IntoTheCloset | Devpost

Inspiration

Many people may not have the physical ability to look through their closet and match together an outfit for the day. This virtual reality environment gives them the freedom and opportunity to search through all the clothes they have uploaded to get a sense of how an outfit would look an a life-sized mannequin before physically getting it from their closet. Or they can upload a picture of themselves and see what the outfit will look like on a real-size version of themselves!

What it does

Uploads any picture of an article of clothing, crops the background, and imports it into a virtual reality, 3d-rendered, interactive environment. Also applies machine learning techniques to classify clothing by properties like garment type and color.

Challenges we ran into

Learning curve on all the platforms we didn't initially know how to use (all of them).

Accomplishments that we're proud of

Using Matlab and openCV to create a transparent background of an image of clothing so users can upload their personal closet into the virtual reality environment.

Using Google Cloud Vision to determine properties of clothing.

Using the HTC Vive to upload images of clothing into an interactive virtual closet which allows you to scroll through your unique closet. You can also resize, rotate and create a duplicate of an article of clothing before placing it on the life sized mannequin in front of you, or the recycle bin behind you!

What we learned

How to use OpenCV on Matlab, how to use google's open source APIs and how to create a virtual, interactive environment on the Vive.

What's next for IntoTheCloset

Upload images from ecommerce platforms or shopping sites to revolutionize online shopping by creating the unique experience of combining your current closet with potential new items.

Implement seamless integration of the three utilized platforms (python, matlab, unity/vive). Create a GUI or app to make photo uploads to the image processing and machine learning algorithms, and into the virtual reality environment autonomous. Autonomously create 3D rendered objects of clothing. Categorize clothing based on assigned tags and suggest outfits based on previous outfits, weather or event attending with the help of Google Vision and Google Prediction.