Our model identifies what animal is being depicted in a drawing from a set of 10 categories!
This project was implemented by Anson Ng and Abigail Wolfensohn as a MAIS202 bootcamp project.
Our model is a convolutional neural network (CNN) that can determine whether a simple doodle of an animal is of a bat, bee, cat, duck, elephant, lion, octopus, rabbit, snail, or whale. At the moment, it has a test accuracy of over 83%.
We trained our model using images from Google's Quick Draw dataset.
We have now also implemented a webapp where users can upload a doodle of an animal and the model will determine a most likely match.