Autoshelf 📚
As part of the NUS Hack&Roll 2020, my team designed an automated cataloguer with the aim of helping non-tech savvy businesses list their items onto online platforms in a seamless manner. The device would be paired with an app that used machine learning to predict the model and make of the product.
Unique Design
Powered by an Arduino, this portable device is meant to accomodate any phone that has our app installed.
The turntable will spin while the phone will take multiple photos of the product. Our future plan was to use the images to stitch them into a 3D model that could be uploaded online for both verification of the product, as well as to showing it off to potential customers.
Since that was a little too advanced, we ended up simply using the photos to feed into our machine learning model to generate a prediction of the item.
The turntable powered by a motor below
Accurate prediction of a shoe
Easy inputting of additional features
Turning the shoe on the turntable