Skip to content

Latest commit

 

History

History
12 lines (7 loc) · 623 Bytes

File metadata and controls

12 lines (7 loc) · 623 Bytes

Image-Classification

This is a classical Binary Image Classification Problem where we need to classify whether the given image belongs to dog or cat.

Data provided: 8000 images ( training set) and 2000 images(test set) of dogs & cats splitted evenly.

We build this model using Convolution Neural Network(CNN).

Libraries Used: keras, theano, tensorflow.

On building the model with single convolution and pooling layer, we achieved an accuracy of 84.52% on training set and 75.10% on test set.

By adding second convolution and pooling layer, we achieved an accuracy of 85.16% on training set and 81.80% on test set.