Google Code-in

Perform Image Classification using Data Augmentation on images using both SwiftCV and External Python Libraries. Compare the results through Loss and Accuracy Graphs.

Task:

1.) Choose a small scale dataset for the task.

2.) Create three individual Image Classification projects.

a.) Don't use Image Augmentation in the first one. Just normal Image Classification.

b.) Use SwiftCV in the second one for Image Augmentation.

c.) Use external python libraries like PIL, numpy etc. for Image Augmentation. You'll need to research a little bit about how Image Augmentations work and what kind of augmentations can be used for your data.

3.) Record various augmentations you tested and implemented.

4.) Record evaluation parameters in form of graphs/numbers like losses, accuracy, average time spent on each epoch, memory required etc.

Present your Colab notebook links as well as your findings in the form of a blog/pdf.