OmniweAR

XR Hackathon Project Write-Up 05/08/22 Project OmniweAR Justin Jasper, Yash Kadadi, Dominik Helmreich, Amita Gondi

Project Objective: The overall objective was to design a program that would create a three dimensional scan of the user’s body and use the scan to model retail clothing on top of the user in a manner that is dimensionally accurate to the user’s body proportions.

Inspiration

Clothing returns for items bought online lose retailers money and hurt the planet. The average online shopping return rate is 30%, and only around 50% of these returns actually end up back in inventory. The rest is discarded in landfills - one truckload a second, to be exact - costing retailers $550B a year and a countless toll on the environment. Over half of the returns come from poor fitting - in fact, according to Medium, “88% of consumers are frustrated with regards to sizing inconsistency.” Retailers and customers alike are in desperate need of a product that allows them to try on clothes they found online before buying and returning them. A few emerging products exist that allow users to see themselves under overlaid images of clothing, but no products currently exist that allow users to accurately predict fitting for retail clothing prior to purchase. These problems inspired us to create Omniwear.

What it does

The minimally viable product that we created uses a 3d scan of one of our founders as its input. The program then models the fit of different articles of clothing on the 3d avatar in AR using a phone camera.

How we built it

We used a third-party scanner app that utilizes photos and LIDAR to recreate a 3D human model. We then import this model into Unity, and superimpose different clothing options onto the model. Then, in a 3D AR environment, the user can look at this fitting and decide whether or not to buy the product.

Challenges we ran into

Learning Unity and AR: We are completely new to XR, and two of our teammates are freshman with relatively less experience. While learning to use Unity, we had a great deal of fun creating our first AR application!

Sourcing and Using Clothing Models: We had a hard time finding clothes that could accurately be rigged to our human model. To fix this, we added individual rigs for each piece of clothing to allow for the fitting and parenting of clothes to the human.

Deploying: For a while, we couldn’t deploy the code to the iPhone. After much research and trial and error, we had to download a dmg of Xcode support for a newer version of iOS in order to build and test the code.

Accomplishments that we're proud of

Clothing models: Each of the pieces of clothing are correctly parented to the base model, meaning if we wanted to strike a specific pose for the human model, the clothes fold and warp realistically around the body.

Switch/Outfit Selector: Our buttons allow different pieces of clothing to be individually swapped out, letting users discover unique varieties and combinations of clothing they can potentially buy.

What we learned

Prior to this project, our members possessed little to no experience with Unity. We learned the necessary mechanisms for modeling and parenting clothing on Unity as well as rigging the clothing items in Blender. We also learned to deploy an AR app to an iOS device.

What's next for OmniweAR

In the future, we plan on incorporating the 3d scanning into our own program. Furthermore, we plan on giving users the ability to use their own body scans to not only model clothing items originally designed in unity (as occurs in our minimally viable product) but also be able to model retail clothing items that they are seeking to purchase. With users’ scans, we would also design a recommendation algorithm for clothes that fit the user’s body well, helping them pick and choose outfits faster. We envision the program not only fitting clothing items onto users’ avatars, but also determining user’s clothing sizings and using an algorithm to give recommendations on sizing based on the user’s preferred fit type (i.e. tight vs slim fit vs baggy clothing). Moreover, the program will also incorporate another recommendation algorithm based on clothing styles to users based on their clothing view-history and past clothing purchases within our program.

After its official launch, our program will exist as a plugin within existing hubs for e-commerce, but the end goal for omniwear is become an online clothing marketplace where users can use their body scans to try on a multitude of apparel items from many different retailers (with which OmniweAR has a partnership), replacing the traditional sense of shopping malls with a virtual alternative.

Built With

• c#
• unity