Business Case:
This project supports the registration process for a citywide dog show by accurately identifying dog breeds using pre-trained image classifiers. By automating breed verification, it helps ensure that only legitimate dog entries are registered, minimizing errors and optimizing event organization.
Technical Solution:
Leveraging Python and pre-trained convolutional neural networks (CNNs), the project utilizes three different CNN architectures—AlexNet, VGG, and ResNet—to classify images as "dogs" or "not dogs" and identify specific dog breeds. The solution involves timing and accuracy assessments for each model, choosing the one best suited to the classification task. The classifier function is used to handle image inputs and return predictions, balancing computational efficiency with classification accuracy.