Amanpradhan/Dogs-vs-Cats

It is a Dogs vs Cats classifier using Convolutional Neural Netwroks

Libraries used :

Keras

Performance metrics: Our accuracy, before augmentation, which was about 50%. After doing data augmentation, it was nearly 75%.

What else can be done to further improve our model :

Use larger dataset - The original dataset was containing 12500 images each, out of which we took only 1000.

Use Fastai Libraries - Fastai has created some libraries which are tailor made for a project like this.