Our problem is to classify the pictures in our dataset to one of the 6 categories as shown in Figure 1.. Our motivation is to find a good machine learning algorithm which could be used to classify different categories of pictures. And this can be used in a wide variety of palaces. And it is a very common demand in daily life.

Dataset is in Folder intel-image-classification/, already splited to training data and test data

File util542 is tool functions for loading training set randomly suffle images, loading and testing models.

You can excute .ipynb files on jupyter notebook, and for .py file, they are running on VSCode with Anaconda Extension (which is really a convinient tool!), or you can copy and paste the cell (begins with #%%) to jupyter for excution

When running each file of to train model, a .h5 file will be generated for this training result.

Some .h5 files are trained model we made, these files really big, some exceeded the file limit 100mb on GitHub. SO we did not contain these file in this repo: XceptonNet Model and resNet model