Multiclass-scene-classification-using-tf.keras

The goal is to classify around 25k images of size 150x150 distributed under 6 categories.
namely {'buildings' -> 0, 'forest' -> 1, 'glacier' -> 2, 'mountain' -> 3, 'sea' -> 4, 'street' -> 5 } using
1)CNN
2)Transfer Learning
DataSource: Intel Image Classification

Usage

python customCNN.py

Or

python transferLearning.py

Dataset

Summary of custom CNN Model.

Accuracy on training and validation Data.

Summary of Model pre-trained on imagenet using Transfer Learning.

Accuracy on training and validation Data.