README.md

metadata

title: VoXMED
emoji: 🐢
colorFrom: red
colorTo: yellow
sdk: streamlit
sdk_version: 1.37.1
app_file: app.py
pinned: false

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

One-Step Respiratory Disease Classifier using Digital Stethoscope Sound - Readme

This project provides a user-friendly Streamlit application to classify respiratory diseases using audio data from a digital stethoscope.

Features:

• Uploads a digital stethoscope audio file (WAV or MP3 format).
• Extracts features from the audio using a pre-trained Audio Set Transfer (AST) model.
• Predicts the most likely respiratory disease based on the extracted features using a deep learning model.
• Displays informative messages and relevant images based on the prediction.

Requirements:

• Python 3.x
• Streamlit (pip install streamlit)
• TensorFlow (pip install tensorflow)
• PyTorch (pip install torch)
• torchaudio (pip install torchaudio)
• transformers (pip install transformers)
• Pillow (pip install Pillow)

Instructions:

1. Download the pre-trained AST model or Import it From the Hugging Face Website and disease classification model:
   • Download the AST model files (e.g., pytorch_model.bin) from https://huggingface.co/MIT/ast-finetuned-audioset-10-10-0.4593 (replace with the actual download URL). Place them in a directory.
   • Download the disease classification model (Model.h5) and place it in the same directory as the AST model files.
2. Update file paths in the code:
   • Unzip the Assets zip file
   • Modify the following paths to reflect your actual locations:
     • 'C:\\Users\\UserName\\Desktop\\RESPIRATORY DISEASE CLASSIFIER\\Model.h5' (path to your disease classification model)
     • 'C:\\Users\\UserName\\Desktop\\RESPIRATORY DISEASE CLASSIFIER\\Assets\\Healthy.gif' (path to the healthy image)
     • 'C:\\Users\\UserName\\Desktop\\RESPIRATORY-DISORDERS-.jpg' (path to the generic respiratory issues image)
     • 'C:\\Users\\UserName\\Desktop\\RESPIRATORY DISEASE CLASSIFIER\\Assets\\COPD.png' (path to the COPD info image )
3. Run the application:
   • Open a terminal and navigate to the directory containing the script (APP.py).
   • Run the script using streamlit run APP.py.
4. Use the application:
   • Upload an audio file from your digital stethoscope.
   • The application will display the predicted disease, relevant information, and images.
   • For COPD prediction, an additional information button can be clicked to display a detailed explanation.

Disclaimer:

This application is for informational purposes only and should not be used for medical diagnosis. Always consult a qualified healthcare professional for any health concerns.