Your task is to perform a binary classification and detect whether there is a crack or not. This version of the project is the simplest and therefore can be awarded with a grade of 3.0 at most.
Besides high accuracy, it is also important to perform all important steps of the neural network creation and training like data preprocessing, architecture selection, results analysis.

Use the following dataset https://www.kaggle.com/datasets/lakshaymiddha/crack-segmentation-dataset. Only the train directory is supposed to be used during the training.