This project is an AI-based emotion detection system that classifies emotions from text input using deep learning.

• Pretrained Deep Learning Model: Uses a Keras-trained model for text-based emotion classification.
• Natural Language Processing (NLP): Tokenizes and processes text for accurate predictions.
• Emotion Prediction: Identifies emotions such as happiness, sadness, anger, love, and more.
• Automated Responses: Generates appropriate responses based on detected emotions.

git clone https://github.com/H-ario-m/mused
cd mused

Ensure you have TensorFlow and required libraries installed:

pip install tensorflow numpy pickle5

To predict emotions, run the model.py script:

python model.py

The script will prompt for user input, predict the emotion, and provide a response.

📁 mused
│── 📜 emotion_detection_model.h5   # Pretrained Keras model
│── 📜 emotion_tokenizer.pickle     # Tokenizer for text preprocessing
│── 📜 label_encoder.pickle         # Label encoder for decoding predictions
│── 📜 load.py                      # Loads the model and makes a sample prediction
│── 📜 model.py                     # Main script for real-time emotion detection
│── 📜 soul.py                      # Basic implementation for testing predictions
│── 📜 README.md                    # Project documentation

1. The model loads a pretrained deep learning model for emotion classification.
2. User input is processed into tokenized sequences using emotion_tokenizer.pickle.
3. The model predicts the emotion based on the input text.
4. The label encoder converts the numerical prediction into a readable label.
5. A predefined response is generated based on the detected emotion.

Input:

You: "I feel so happy today!"

Output:

Predicted Emotion: joy  
Bot: "Joy is such a beautiful feeling! 😄"

• Support for multilingual emotion detection.
• Addition of more diverse emotion responses.
• Integration with chatbot systems for enhanced conversations.