Here, in this example, we’ll create a simple preprocessing and augmentation model.
Dataset for the model
For the dataset, just collect some images from the internet and put it in one folder.
Install the library
Check the install page for installation instructions.
Importing dependencies from Hocrox
Let’s import the dependencies from Hocrox.
Here we make a simple model with some basic preprocessing and augmentation layers.
from hocrox.model import Model from hocrox.layer.preprocessing.transformation import Grayscale, Resize from hocrox.layer.augmentation.flip import RandomFlip from hocrox.layer.augmentation.transformation import RandomRotate from hocrox.layer import Read, Save
Making the model
The model class provides an easy
.add() method to add layers. We will use the .add() method here to add some layers.
# Initializing the model model = Model() # Adding model layers model.add(Read(path="./img")) model.add(Resize((224, 224))) model.add(Grayscale()) model.add(RandomFlip(probability=1.0, number_of_outputs=2)) model.add(RandomRotate(probability=1.0, start_angle=-10, end_angle=10, number_of_outputs=5)) model.add(Save("./processed_images"))
Summary of the model
Once we defined the model, it is a good idea to print the model summary to make sure the pipeline is correct.
To print summary, we have a simple
.summary() method. We will use it here.
# Printing the summary of the model print(model.summary())
Transforming the images
Once we are done with the model pipeline, we can use the
.transform() method transform the images based on the model pipeline.
The transform method transform the images and saves it into the given path.
# Apply transform to the images model.transform()