Cough_Audio_Detection_ML

A web app that detects cough audio and displays the classification results in Streamlit.

Project Overview

This project is a real-time cough detection system built using Python, Streamlit, and Machine Learning. It records audio via a microphone, extracts MFCC features, and classifies the sound as cough or non-cough using a Random Forest classifier.

Features

• Real-time & Manual Recording – Users can record live audio or manually upload files.
• Feature Extraction – Extracts MFCCs & Spectrograms from audio.
• Machine Learning Classification – Detects cough events using Random Forest Classifier.
• Data Visualization – Displays waveform, MFCCs, and classification results.
• Live Monitoring – Continuously detects coughs and logs them.
• Cough History & Alerts – Tracks cough trends and alerts users when multiple coughs are detected.

---

Installation & Setup

1. Clone the Repository

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

2. Install Dependencies

Create a virtual environment (recommended) and install required packages:

python3 -m venv venv
source venv/bin/activate  # On Mac/Linux
venv\Scripts\activate  # On Windows
pip install -r requirements.txt

3. Run the Application

streamlit run app.py

---

Usage Guide

1. Running the App

After launching the app, select a mode:

• Live Monitoring: Real-time audio detection.
• Manual Recording: Record & analyze short audio clips.

2. Uploading and Classifying Audio

• Click "Upload File" to test existing .wav files.
• View waveform, MFCCs, and spectrogram.
• Model will classify audio as Cough or Non-Cough.

3. Training the Model

• Upload cough and non-cough audio samples.
• Extracted MFCCs are used to train a Random Forest model.
• Model is saved as Random_forrest.pkl for future use.

4. Real-time Cough Monitoring

• Click "Start Live Monitoring" to continuously detect coughs.
• If 10 coughs are detected within a short period, an alert is displayed.
• View cough event history & trend graphs.

---

Project Structure

Cough_Detection_ML/
│── app7.py                 # Main Streamlit App
│── Random_forrest.pkl     # Pre-trained ML Model
│── requirements.txt       # Dependencies
│── data/
│   ├── cough_1.wav       # Example cough audio
│   ├── speech_1.wav      # Example non-cough audio
│── utils/
│   ├── feature_extraction.py  # MFCC & Spectrogram Functions
│   ├── audio_processing.py    # Audio Preprocessing
│   ├── model_training.py      # Machine Learning Model Training

---

Technologies Used

• Programming: Python 3.13.1
• Libraries: Streamlit, SoundDevice, Librosa, Scikit-Learn, Pandas, Matplotlib
• Machine Learning: Random Forest Classifier

---

Future Improvements

Planned Features:

• IoT Integration (e.g., use with smartwatches, medical devices).
• Cloud Deployment (Google Cloud, AWS, Azure).
• CNN Deep Learning Model for improved accuracy.
• Mobile App Version (Android/iOS).

---

Contact

Email: akaqinlang@gmail.com GitHub: Scott2001404