Why Flow2Ml?
Flow2ML is an open source library to make machine learning process much simpler. It loads the image data and applies the given filters and returns train data, train labels, validation data and validation labels.
For all these steps it just take 3 lines of code. It mostly helps beginners in the field of machine learning and deep learning where the user would deal with image related data.
Installation
Install Flow2ML python files via pip.
pip install flow2ml==1.0.2
Write pre-Processing steps in minutes.
import flow2ml as f2m
# To be given input by the user.
img_dimensions = (299,299)
test_val_split = 0.1
flow = f2m.Flow( 'dataset_dir' , 'data_dir' )
filters = ["median","laplacian","gaussian"]
flow.applyFilters( filters )
(train_x, train_y, val_x, val_y) = flow.getDataset( img_dimensions, test_val_split )
# To be given input by the user.
img_dimensions = (299,299)
test_val_split = 0.1
flow = f2m.Flow( 'dataset_dir' , 'data_dir' )
filters = ["median","laplacian","gaussian"]
flow.applyFilters( filters )
(train_x, train_y, val_x, val_y) = flow.getDataset( img_dimensions, test_val_split )