Applying Python tools to use in an image classifier to identify Dog breeds
A hypothethical city is hosting a citywide dog show and I have volunteered to help the organizing committee with contestant registration. Every participant that registers must submit an image of their dog along with biographical information about their dog. The registration system tags the images based upon the biographical information. Some people are planning on registering pets that aren’t actual dogs. Apply Python tools to use a image classifier to identify dog breeds.
Tasks: Using Python skills, to determine which image classification algorithm works the "best" on classifying images as "dogs" or "not dogs". Determine how well the "best" classification algorithm works on correctly identifying a dog's breed. The Input is an image. The output determines what the image depicts. (for example, a dog). Be mindful of the fact that image classifiers do not always categorize the images correctly. Time how long each algorithm takes to solve the classification problem. With computational tasks, there is often a trade-off between accuracy and runtime. The more accurate an algorithm, the higher the likelihood that it will take more time to run and use more computational resources to run.
We'll use a CNN that has already learned the features from a giant dataset of 1.2 million images called ImageNet(opens in a new tab). There are different types of CNNs that have different structures (architectures) that work better or worse depending on the criteria. With this project, we'll explore the three different architectures (AlexNet, VGG, and ResNet) and determine which is best for our application.
The files that I have worked on - get_input_args.py, get_pet_labels.py, classify_images.py, adjust_results4_isadog.py, calculates_results_stats.py, print_results.py