Cat Emotion Classifier

This is a Streamlit application that classifies the emotions of cats from uploaded images using a Vision Transformer (ViT) model.

You can try the application here.

Examples

Features

• Upload an image of a cat and classify its emotion.
• Display the uploaded image.
• Show the classification results as a horizontal bar chart with confidence scores for each class.

Installation

1. Clone the repository:

   git clone https://github.com/semihdervis/cat-emotion-classifier.git
   cd cat-emotion-classifier

2. Create a virtual environment and activate it:

   python -m venv venv
   source venv/bin/activate  # On Windows, use `venv\Scripts\activate`

3. Install the required dependencies:

   pip install -r requirements.txt

Usage

1. Run the Streamlit app:

   streamlit run app.py

2. Open your web browser and go to http://localhost:8501.

3. Upload an image of a cat and click the "Classify Image" button to see the classification results.

   Note: The first prediction may take some time as the Vision Transformer (ViT) model needs to be loaded from Hugging Face.

Model

The application uses a Vision Transformer (ViT) model fine-tuned for cat emotion classification, available on Hugging Face as semih-dervis/cat-emotion-classifier.

License

This project is licensed under the MIT License. See the LICENSE file for details.