I started a new pet project. The main idea is to create an application that involves children in the drawing process, recognizing what they have drawn.
The MVP uses the “Quick, Draw!” data set and Create ML tool.
Source code for iOS application is available on GitHub
Result: MVP works fine and recognizes many animals.
Next steps: make ML model smarter, add game characters which will show their answers.
I think it will be funny to have 2 characters:
1 - A fox. It tries to recognize an animal (cat, dog, …)
2 - A robot. It tries to recognize a thing (car, airplane, …)
