🧠 Human Mental Health Condition Detection

A real-time facial emotion detection app using deep learning. This project can be a prototype for detecting mental health conditions based on facial expressions.

---

📸 Features

• ✅ Upload an image and detect emotional expressions
• ✅ Recognize seven key emotions: Angry, Disgust, Fear, Happy, Sad, Surprise, Neutral
• ✅ Save output images with bounding boxes and labels
• ✅ Clean UI using modern Streamlit design
• ✅ Uses Haar cascade for fast face detection
• ✅ Lightweight and responsive

---

🚀 Demo

🌐 Try Now: https://share.streamlit.io/ 📦 Download the dataset used: FER2013

---

💻 Tech Stack

Technology	Use Case
Python	Core programming language
Streamlit	UI rendering and deployment
OpenCV	Face detection
Keras/TensorFlow	Emotion classification model
Pillow	Image processing and saving

---

🗂️ Folder Structure

├── model.h5 # Trained deep learning model
├── haarcascade_frontalface_default.xml# Haar cascade classifier
├── emotion.py # Main Streamlit app
├── requirements.txt # List of dependencies
└── README.md # This file

---

⚙️ Setup Instructions

1. Clone the Repository

git clone [https://github.com/your-username/human-mental-health-detection.git](https://github.com/GHOSH2341/Human-Mental-Health-Condition-Detection.git)
cd human-mental-health-detection

2. Create Virtual Environment

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install Dependencies

pip install -r requirements.txt

4. Add Model & Cascade Files

• Place model.h5 (trained Keras model) in the root directory.

• Place haarcascade_frontalface_default.xml in the root directory.

5. Run the App

streamlit run emotion.py

🌍 Deployment Options

• 🟢 Streamlit Cloud
• Push code to a public GitHub repository
• Log in to Streamlit Cloud
• Deploy the app from your repository
• Add model and XML files as static assets or use environment variables/secrets

🟡 Hugging Face Spaces (Optional)

• Create a new Gradio or Streamlit Space
• Upload files and paste your emotion.py

🪪 License

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