Update app.py

app.py

@@ -11,17 +11,19 @@ def process_video_audio(video_path, audio_path):
 
     wav = pt.tensor(list(audio_path[1]))
     
-    train_visual = np.zeros([1, 120, 120, 3, 10])
-    train_audio_wave = np.zeros([1, 261540])
-    train_audio_cnn = np.zeros([1, 150, 512, 1])
+    train_visual = np.zeros([2, 120, 120, 3, 10])
+    train_audio_wave = np.zeros([2, 261540])
+    train_audio_cnn = np.zeros([2, 150, 512, 1])
     
     mfcc = torchaudio.transforms.MFCC(n_mfcc=150, melkwargs={"n_fft": 1022, "n_mels": 150})
     
     face_cascade = cv2.CascadeClassifier(cv2.data.haarcascades + 'haarcascade_frontalface_default.xml')
 
     if len(wav) > 261540:
+        print(wav.shape)
         train_audio_wave[0, :] = wav[:261540]
     else:
+        print(wave.shape)
         train_audio_wave[0, :len(wav)] = wav[:]
     train_audio_cnn[0, :, :, 0] = mfcc(train_audio_wave[0])