| # Emotion Recognition Model |
|
|
| This repository hosts a custom-tuned Emotion Recognition model designed to analyze facial expressions in real-time video streams. The model predicts seven distinct emotions: **Angry**, **Disgust**, **Fear**, **Happy**, **Neutral**, **Sad**, and **Surprise**, and calculates a positivity level based on the predicted emotions. |
|
|
| ## Model Details |
|
|
| - **Model Architecture:** The model is based on a fine-tuned VGG16 architecture using TensorFlow. |
| - **Input Size:** 224x224 pixels |
| - **Emotion Labels:** |
| - Angry |
| - Disgust |
| - Fear |
| - Happy |
| - Neutral |
| - Sad |
| - Surprise |
|
|
| ## How to Use |
|
|
| 1. Clone the repository and install the dependencies: |
|
|
| ```bash |
| pip install -r requirements.txt |
| ``` |
|
|
| 2. Load the model and run predictions on images: |
|
|
| ```python |
| from tensorflow.keras.models import load_model |
| import cv2 |
| import numpy as np |
| from tensorflow.keras.preprocessing.image import img_to_array |
| from tensorflow.keras.applications.vgg16 import preprocess_input |
|
|
| model = load_model('emotion_model_finetuned224.h5') |
|
|
| # Example usage with an image |
| image = cv2.imread('example.jpg') |
| face = cv2.resize(image, (224, 224)) |
| face = img_to_array(face) |
| face = preprocess_input(face) |
| face = np.expand_dims(face, axis=0) |
| predictions = model.predict(face) |
| print(predictions) |
| ``` |
|
|
| ## Gradio Interface (Optional) |
|
|
| If you want to deploy the model using Gradio, you can use the `app.py` file provided: |
|
|
| ```bash |
| pip install gradio |
| python app.py |
| ``` |
|
|
| Visit the Gradio interface at `http://localhost:7860` and upload an image to see the predicted emotion. |
|
|
| ## Dependencies |
|
|
| - `tensorflow` |
| - `opencv-python` |
| - `numpy` |
| - `requests` |
| - `gradio` (for optional web interface) |
|
|
| ## License |
|
|
| This project is licensed under the MIT License. Feel free to use and modify as needed. |
|
|
| ## Contact |
|
|
| For questions or support, please open an issue on the GitHub repository or contact Ahmed directly at [migdady@gmail.com](mailto:migdady@gmail.com). |
|
|
|
|
