/CatDogClassification

A CNN to distinguish between cat images and dog images with a fairly high accuracy with less data

Primary LanguageJupyter Notebook

Cats VS Dogs

A CNN to distinguish between cats and dogs.

Training set of 2000 images, 1000 images of cats and dogs each.

Accuracies

100 epochs

With Augmentation

Gives a training accuracy of 89% and a validation accuracy of 87.1%

Without Augmentation

Give a training accuracy of 99% and validation accuracy of 76%

50 epochs

Without Augmentation

Gives a training accuracy of 85% and validation accuracy of 84%

Trained on Google Colab

Code

Cats_vs_Dogs_model2.ipynb (50 epochs)