The final app will consist of modules to prepare the dataset, train the models, and run an inference with the models using input from your camera. This will require the following components:
- main.py: This is the main script for starting the application and localizing the head (of the pets) in real time.
- data.py: This is a module to download and prepare the dataset for training.
- classification.py: This is a script to train a classifier network.
- localization.py: This is a script to train and save a localization network.
After preparing the dataset for training, we will do the following to complete our app:
- We will first train a classification network using transfer learning.
- Next, we will train an object localization network, again using transfer learning.
- After we create and train our localization network, we will run our main.py script to localize the heads...
