Waldo

Inspiration

The constant struggle of losing keys, wallets, T.V. remotes, and more when at home. We wanted to build Waldo to save the world (and ourselves) from lost time and frustration.

What it does

Waldo detects general items from a predefined object library or specific items from a library personalized by the user. When the user loses something, all they have to do is search for it in our app and Waldo will present them with the last image in which the last item was detected. This can help jar the user's memory, narrow down a range of search, and find things that are hiding in plain sight.

How we built it

Waldo was built on top of the YOLO object detection model with the help of Scikit Learn and PyTorch. A mobile interface for the system was built using Android Studio. Design elements were created using Adobe XD and Adobe Illustrator.

Challenges we ran into

Dealing with multiple thread collisions while running the models on our personal server. These issues prevented us from testing the functionality of the interface for quite some time.