Apppublic

edyfy/Depression-Detection-Voice

sourceHugging Facemitupdated 1y agoView on Hugging Face

0likes

App README

Depression Detection (Voice)

This Space uses a voice input to detect signs of depression using a trained ML model.

🎙️ Depression Detection from Voice

This project is uses machine learning to predict the risk of depression from short audio samples of speech. ⚠️ Disclaimer: This tool is for educational purposes only and not for clinical or medical use.

📌 Features

—Upload or record voice samples directly in the app.
—Extracts audio features (e.g., MFCCs) using librosa.
—Predicts probability of depression (0 = low, 1 = high) using a trained model.
—Interactive Gradio UI for testing and demonstration.

🛠️ Installation

1. Clone the repository

bash

git clone https://github.com/your-username/dep_voice_scaffold.git
cd dep_voice_scaffold

2. Set up environment

Install dependencies with pip:

bash

pip install -r requirements.txt

Or if you prefer conda:

bash

conda create -n depvoice python=3.10
conda activate depvoice
pip install -r requirements.txt

▶️ Running the App

Run the Gradio app:

bash

python app.py

This will launch a local web interface (usually at http://127.0.0.1:7860/).

📂 Project Structure

dep_voice_scaffold/
│── app.py               # Main Gradio app
│── requirements.txt      # Dependencies
│── configs/
│    └── default.yaml     # Feature extraction & model config
│── artifacts/
│    ├── model.pkl        # Trained model
│    └── metrics.yaml     # Evaluation metrics (with threshold)
│── src/
│    └── depvoice/        # Feature extraction code

📊 Example Output

After uploading a voice file, the app will return:

—Probability scores (Depressed vs Not Depressed)
—Final prediction (based on threshold in metrics.yaml)

⚠️ Disclaimer

This project is for academic purposes only. It is not a diagnostic tool and must not be used in clinical settings.

✨ Made with Python, Gradio, and ❤️ for research.