update app.py

app.py

@@ -138,7 +138,7 @@ def inv_scaled_ou(matrix_spec):
     return matrix_spec
 
 
-def prediction(weights_path, name_model, audio_dir_prediction, dir_save_prediction, audio_input_prediction,
+def prediction(weights_path, name_model, audioData, dir_save_prediction, audio_input_prediction,
 audio_output_prediction, sample_rate, min_duration, frame_length, hop_length_frame, n_fft, hop_length_fft):
     """ This function takes as input pretrained weights, noisy voice sound to denoise, predict
     the denoise sound and save it to disk.
@@ -154,9 +154,9 @@ audio_output_prediction, sample_rate, min_duration, frame_length, hop_length_fra
     print("Loaded model from disk")
 
     # Extracting noise and voice from folder and convert to numpy
-    audio = audio_files_to_numpy(audio_dir_prediction, audio_input_prediction, sample_rate,
-                                 frame_length, hop_length_frame, min_duration)
-
+    # audio = audio_files_to_numpy(audio_dir_prediction, audio_input_prediction, sample_rate,
+                                #  frame_length, hop_length_frame, min_duration)
+    audio = audioData
     #Dimensions of squared spectrogram
     dim_square_spec = int(n_fft / 2) + 1
     print(dim_square_spec)
@@ -191,8 +191,8 @@ audio_output_prediction, sample_rate, min_duration, frame_length, hop_length_fra
 
 def denoise_audio(audioName):
   testNo = audioName
-  audio_dir_prediction = os.path.abspath("/")+ str(testNo) +".wav"
-  sample_rate, data = wavfile.read(audio_dir_prediction)
+#   audio_dir_prediction = os.path.abspath("/")+ str(testNo) +".wav"
+  sample_rate, data = audioName[0], audioName[1]
   len_data = len(data)  # holds length of the numpy array
 
   
@@ -204,7 +204,7 @@ def denoise_audio(audioName):
   dir_save_prediction = os.path.abspath("/")
   audio_output_prediction = "test"+ testNo+".wav"
   audio_input_prediction = [testNo +".wav"]
-  sample_rate = 8000
+
   min_duration = t
   frame_length = 8064
   hop_length_frame = 8064
@@ -213,10 +213,10 @@ def denoise_audio(audioName):
 
   dim_square_spec = int(n_fft / 2) + 1
 
-  prediction(weights_path, name_model, audio_dir_prediction, dir_save_prediction, audio_input_prediction,
+  prediction(weights_path, name_model, data, dir_save_prediction, audio_input_prediction,
           audio_output_prediction, sample_rate, min_duration, frame_length, hop_length_frame, n_fft, hop_length_fft)
   print(audio_output_prediction)
-  return audio_output_prediction
+  return audio_output_prediction[0]
 
 
 examples = [