Cat vs Dog Classifier 🐕‍🦺🐈

This is a simple cat vs dog classifier that uses a Support Vector Machine (SVM) to classify images of cats and dogs. The model is trained on a dataset of 25,000 images of cats and dogs, and the accuracy of the model on the testing data is 90%.

How to run the project

1. Clone the repository to your local machine.
2. Install the required dependencies by running pip install -r requirements.txt.
3. Download the dataset from Kaggle and extract it to the data directory.
4. Run the main.py script to preprocess the data, train the model, and start the GUI.

Project structure

• data/: Directory containing the dataset of cat and dog images.
  • train/: Directory containing the training images.
  • test/: Directory containing the testing images.
• src/: Directory containing the source code.
  • preprocess.py: Module for preprocessing the images.
  • train_model.py: Module for training the SVM classifier.
  • predict.py: Module for making predictions on new images.
  • gui.py: Module for creating the Gradio-based GUI.
• models/: Directory containing the trained SVM model.
• main.py: Main script that calls the functions in the src modules to preprocess the data, train the model, and start the GUI.
• requirements.txt: List of required Python packages.
• README.md: This file.

Results

The accuracy of the model on the testing data is 90%.

Acknowledgements

This project is Task 3 of the Prodigy InfoTech ML internship.