Designed and implemented a Convolutional Neural Network to identify and classify 60,000 32x32 images.
Dataset: CIFAR-10
Below is an image demonstrating the 10 different classes that the network was able to identify:
We achieved accuracy scores of 91% for the training data and an 80% overall testing accuracy.
After creating an accurate model, we then implemented an interactive app that allowed users to classify their own images.
Below is a screenshot of the network correctly classifying my team member's dog:
