Practice Session 2

The goal of this project is to build a cat family classification model using convolutional neural networks (CNN) and the Fast.ai library. This model will be able to accurately classify between lion, tiger, cougar, caracal, cheetah and ocelot images. We will use transfer learning with a pre-trained ResNet34 architecture as our base for training the model on our dataset of cats from each species in order to achieve high accuracy results quickly and efficiently.

Video

Objectives

1. Install and Import Required libraries
2. How Make Your Own Data
3. Verify the Data
4. Make Data Loaders
5. Make Learner
6. Train the Model
7. Interpretation

Highlights

Following were the highlight’s:

1. Remaking walkwithfastai course ( https://walkwithfastai.com/ )
2. Lesson One Part 2: Cat Family Images classification

Resources :

1. Slides Notes Python Notebook
2. Original Lesson
3. Amazon Book
4. Free Book Version

Acknowledgements

Special Thanks to @Jeremy Howard