Welcome to my Dog Breed Classifier project! This project showcases my skills in deep learning, transfer learning, and cloud deployment. My goal was to build a reliable dog breed classifier that can identify over 120 different dog breeds with high accuracy.
- Achieved an impressive accuracy of over 93% in classifying over 120 dog breeds.
- Utilized a dataset containing more than 20,500 dog images.
- Deployed the model on Hugging Face Spaces for easy access and testing.
Try it out: Check out the live demo here.
The foundation of this project is the Stanford Dogs Dataset, a comprehensive collection of dog images. You can learn more about the dataset by visiting the Stanford Dogs Dataset Reference.
This project leverages various libraries and tools to achieve its goals:
- TensorFlow: Deep learning framework for building and training the dog breed classifier.
- NumPy: For numerical operations and data manipulation.
- Pandas: Used for data analysis and postprocessing.
- Scikit-learn: Applied for model evaluation and metrics.
- Matplotlib: Utilized for data visualization and result presentation.
- Gradio: Integrated for creating a user-friendly web interface for testing the model.
-
Download the Jupyter notebook provided in this repository to access functions for programmatically downloading the original dataset.
-
Follow the instructions in the notebook to set up the project and train your own dog breed classifier.
Contributions are welcome! Please open a pull request if you have any improvements or new features to add.
This project is licensed under the MIT License.