This project develops an animal image classification system leveraging the ResNet50 model pre-trained on ImageNet. The classifier distinguishes among three animal categories: cats, dogs, and snakes, with a demonstrated high accuracy on a balanced dataset.
- Utilizes the ResNet50 model pre-trained on ImageNet for high accuracy.
- Classifies images into three categories: cats, dogs, and snakes.
- Includes a custom-built GUI using Tkinter for interactive classification.
- Employs data augmentation techniques to enhance model generalization.
-
Clone the repository: git clone https://github.com/AlvaroVasquezAI/Animal_Image_Classification.git
-
Navigate to the project directory:
-
Install the required dependencies: pip install -r requirements.txt
To run the classifier with the GUI, execute (Make sure you are in the correct directory): python main.py
Follow the GUI prompts to select and classify images.
The dataset consists of 3,000 images, evenly distributed among the three categories. Images were resized to 256x256 pixels to facilitate processing.
The project follows a systematic approach:
- Data preprocessing and augmentation to prepare the dataset.
- Loading and fine-tuning the ResNet50 model.
- Training the model and evaluating its performance.
- Developing a Tkinter-based GUI for real-time classification.
The model achieved an accuracy of approximately 98.67% on the test dataset, showcasing the effectiveness of transfer learning in image classification.
Predictions:
- Expand the dataset to include more animal classes.
- Explore advanced deep learning models to enhance classification accuracy.
- Develop a web or mobile application to increase accessibility.
- Álvaro García Vásquez