app.py

import gradio as gr
import torchaudio
from speechbrain.inference.enhancement import WaveformEnhancement
import torch

# Load the SpeechBrain enhancement model
enhance_model = WaveformEnhancement.from_hparams(
    source="speechbrain/mtl-mimic-voicebank",
    savedir="pretrained_models/mtl-mimic-voicebank",
)

def enhance_audio(input_audio):
    # Load the uploaded audio file
    waveform, sample_rate = torchaudio.load(input_audio)
    
    # Enhance the audio
    enhanced_waveform = enhance_model.enhance_batch(waveform)
    
    # Save the enhanced audio to a file
    output_path = "enhanced_audio.wav"
    torchaudio.save(output_path, enhanced_waveform.cpu(), sample_rate)
    
    return output_path

# Set up the Gradio interface
demo = gr.Interface(
    fn=enhance_audio,
    inputs=gr.Audio(type="filepath"),  # Upload an audio file
    outputs=gr.Audio(type="filepath"),  # Download the enhanced audio
)

demo.launch()