/tensorflow_imagenet

Primary LanguageJupyter Notebook

Transfer Learning for Image Classification

This repository contains code for transfer learning on three different image classification tasks: malaria classification, Stanford Online Products classification, and bunny classification. All three tasks use VGG19 as the base model for transfer learning. The code is implemented in Python and uses TensorFlow Core for training.

Files

The repository contains the following files:

  • malaria.py: Python script for training the VGG19 model on a malaria image dataset.
  • stanford.py: Python script for training the VGG19 model on the Stanford Online Products dataset.
  • bunny.py: Python script for training the VGG19 model on a bunny image dataset.
  • malaria_core.ipynb: Jupyter notebook containing code for training and evaluating the malaria classification model.
  • stanford_core.ipynb: Jupyter notebook containing code for training and evaluating the Stanford Online Products classification model.
  • bunny_core.ipynb: Jupyter notebook containing code for training and evaluating the bunny classification model.

Usage

To install the necessary libraries.

python -m pip install -r requirements.txt

To use this code, you will need to have TensorFlow and tensorflow-dataset installed. You can then run the Python scripts for training the models or open the Jupyter notebooks for a more interactive experience.

python bunny.py
python stanford.py
python malaria.py

Acknowledgements

The malaria, bunny and Stanford Online Products datasets were obtained from tensorflow offical website.

License

This project is licensed under the MIT License. See the LICENSE file for details.