/ResNet-Intel-Image-Classification

Building a classifier on intel image dataset using transfer learning techniques and ResNet model

Primary LanguageJupyter Notebook

ResNet Intel Image Classification

Building a ResNet model on intel image dataset

Dataset on kaggle

Contents

Data Augmentation

Here we use image data generator for train and test data.

There are 14034 images belonging to 6 classes for train data and 3000 images for test data

Modeling

Here we use a pre-trained model.

We use ResNet50 model with ImageNet weights and remove top of the model.

We add a dense layer with 6 units for the output layer

Evaluation

  • Accuracy and loss plotting

  • p

Prediction

Predicting an image using our model