import gradio as gr | |
from speechbrain.pretrained import MetricGAN | |
# Load the MetricGAN model from Hugging Face | |
model = MetricGAN.from_hparams(source="speechbrain/metricgan-plus-voicebank") | |
# Define a function to enhance speech | |
def enhance_speech(audio_path): | |
# Process the uploaded audio file through the model | |
enhanced_audio = model.enhance(audio_path) | |
return enhanced_audio | |
# Set up the Gradio interface | |
iface = gr.Interface( | |
fn=enhance_speech, | |
inputs=gr.Audio(source="upload", type="filepath"), | |
outputs=gr.Audio(type="filepath"), | |
title="Speech Enhancement" | |
) | |
# Launch the Gradio interface | |
iface.launch() | |