I have to make a CNN model in Jupyter notebook. I have been given a zip file which has different folders which are names class_a, class_b etc. with images in it. It also have a labels.csv file in it which has the image path and class_label.

How do I import this data into my model for it to read the images and work. I’m not quite sure how to start this assignment. This is the criteria of the assignment:

Uncompress dataset to:

Read the dataset images.
Transform them accordingly in preparation for modelling.
Do image classification on the test dataset.
Among many parameters, hyperparameters, and data preparation steps you can do to use with CNN, explain why you took the approach “A” instead of “B”.
Your goal ultimately is to classify an image between 5 possible classes using a CNN model.

This dataset is anonymised, having 500 samples of 5 classes, totalling 2500 image files.

An accuracy of 20% means your model is randomly guessing (5 classes, 20% probability each).

For timing planning, a model running on 8CPU should take around 10 seconds per epoch.

Hence, this dataset is made to run on CPU, as a worst case scenario, it should take 1min per epoch.

As advised during tutoring, it is expected that you will create a CNN from the ground up and can(albeit not needed) also use transfer learning as an additional experiment.

Please help ????

I have tried unzipping the data in the notebook and I also have added the path to the cvs file but it keeps coming up with file note found and even when it finds the cvs file the images are not being able to be found