mozayed007 / pre-trained-image-classifier

Image Classification for a City Dog Show using pre-trained Classifiers

Dog Breeds Images Classifier

This is a pre-trained dog breed image classifier. It uses different architectures like Resnet, Alexnet, and VGG for classification. The project is part of the AWS-sponsored Scholarship: AI Programming with Python Nanodegree on Udacity.

Prerequisites

Make sure you have Python installed on your system. You can download Python from here.

Requirements

This project requires the following Python libraries:

• ast
• PIL (Python Imaging Library)
• torchvision
• torch

You can install these libraries using pip:

pip install ast
pip install pillow
pip install torchvision
pip install torch

Usage

You can use the classifier with the following commands:

For Resnet architecture:

python check_images.py --dir pet_images/ --arch resnet  --dogfile dognames.txt > resnet_pet-images.txt

For Alexnet architecture:

python check_images.py --dir pet_images/ --arch alexnet --dogfile dognames.txt > alexnet_pet-images.txt

For VGG architecture:

python check_images.py --dir pet_images/ --arch vgg  --dogfile dognames.txt > vgg_pet-images.txt

Parameters

• --dir: Directory of the pet images.
• --arch: The architecture to be used for the classifier. It can be resnet, alexnet, or vgg.
• --dogfile: The file containing the names of the dog breeds.

The output of the classifier will be saved in a text file.

---

If you need more specific information or have any other requirements, please let me know.