Update app.py

app.py

@@ -1,6 +1,7 @@
 import gradio as gr
 import numpy as np
 import soundfile as sf
+import noisereduce as nr
 import spaces
 import torch
 import torchaudio
@@ -323,16 +324,20 @@ def voice_conversion(input, reference, steps, guidance, speed):
     print("[INFO] | Final audio normalized.")
 
     # Smoothen the audio to reduce distorted audio
-    def butter_lowpass_filter(data, cutoff=3000, fs=sr_current, order=6):
-        
+    def butter_bandpass_filter_filtfilt(data, lowcut=80, highcut=6000, fs=sr_current, order=4):
         nyq = 0.5 * fs
-        normal_cutoff = cutoff / nyq
-        b, a = butter(order, normal_cutoff, btype='low', analog=False)
-        y = lfilter(b, a, data)
+        low = lowcut / nyq
+        high = highcut / nyq
+        b, a = butter(order, [low, high], btype='band')
+        y = filtfilt(b, a, data)
         return y
 
-    final_audio = butter_lowpass_filter(final_audio)
+    final_audio = butter_bandpass_filter_filtfilt(final_audio)
     print("[INFO] | Final audio smoothed with low-pass filter.")
+
+    noise_profile = nr.get_noise_profile(final_audio, sr_current)
+    final_audio = nr.reduce_noise(y=final_audio, sr=sr_current, y_noise=noise_profile, prop_decrease=1.0)
+    print("[INFO] | Final audio noise reduced using noisereduce.")
     
     # Save the audio to a temporary WAV file
     print("[INFO] | Saving final audio to a temporary WAV file.")