Update app.py

app.py

@@ -34,33 +34,25 @@ def one_shot(image,input_text,gender):
           sound = AudioSegment.from_file(f.name, format="mp3")
           sound.export("/content/audio.wav", format="wav")
      waveform, sample_rate = torchaudio.load("/content/audio.wav")
-       
-     torchaudio.save("/content/audio.wav", waveform, sample_rate, encoding="PCM_S", bits_per_sample=16)
-     image = Image.open(image_in)
-     image = pad_image(image)
-     image.save("/content/image_pre.png")
-     pocketsphinx_run = subprocess.run(['pocketsphinx', '-phone_align', 'yes', 'single', '/content/audio.wav'], check=True, capture_output=True)
-     jq_run = subprocess.run(['jq', '[.w[]|{word: (.t | ascii_upcase | sub("<S>"; "sil") | sub("<SIL>"; "sil") | sub("\\\(2\\\)"; "") | sub("\\\(3\\\)"; "") | sub("\\\(4\\\)"; "") | sub("\\\[SPEECH\\\]"; "SIL") | sub("\\\[NOISE\\\]"; "SIL")), phones: [.w[]|{ph: .t | sub("\\\+SPN\\\+"; "SIL") | sub("\\\+NSN\\\+"; "SIL"), bg: (.b*100)|floor, ed: (.b*100+.d*100)|floor}]}]'], input=pocketsphinx_run.stdout, capture_output=True)
-     with open("test.json", "w") as f:
-         f.write(jq_run.stdout.decode('utf-8').strip())
-    
-     os.system(f"cd /content/one-shot-talking-face && python3 -B test_script.py --img_path /content/image_pre.png --audio_path /content/audio.wav --phoneme_path /content/test.json --save_dir /content/train")
+
+     try:
+         torchaudio.save("/content/audio.wav", waveform, sample_rate, encoding="PCM_S", bits_per_sample=16)
+         image = Image.open(image_in)
+         image = pad_image(image)
+         image.save("/content/image_pre.png")
+         pocketsphinx_run = subprocess.run(['pocketsphinx', '-phone_align', 'yes', 'single', '/content/audio.wav'], check=True, capture_output=True)
+         jq_run = subprocess.run(['jq', '[.w[]|{word: (.t | ascii_upcase | sub("<S>"; "sil") | sub("<SIL>"; "sil") | sub("\\\(2\\\)"; "") | sub("\\\(3\\\)"; "") | sub("\\\(4\\\)"; "") | sub("\\\[SPEECH\\\]"; "SIL") | sub("\\\[NOISE\\\]"; "SIL")), phones: [.w[]|{ph: .t | sub("\\\+SPN\\\+"; "SIL") | sub("\\\+NSN\\\+"; "SIL"), bg: (.b*100)|floor, ed: (.b*100+.d*100)|floor}]}]'], input=pocketsphinx_run.stdout, capture_output=True)
+         with open("test.json", "w") as f:
+             f.write(jq_run.stdout.decode('utf-8').strip())
+        
+         os.system(f"cd /content/one-shot-talking-face && python3 -B test_script.py --img_path /content/image_pre.png --audio_path /content/audio.wav --phoneme_path /content/test.json --save_dir /content/train")
+         
+     except Exception as e:
+         print(e)
      return "/content/train/image_pre_audio.mp4"
      exit()
        
-      #1. Extract the frames from the video file using PyVideoFramesExtractor
-     os.system(f"python /content/PyVideoFramesExtractor/extract.py --video=/content/train/image_pre_audio.mp4")
-    
-      #2. Improve image quality using GFPGAN on each frames
-     os.system(f"python /content/GFPGAN/inference_gfpgan.py --upscale 2 -i /content/extracted_frames/ -o /content/video_results  --bg_upsampler realesrgan")
-    
-      #3. Merge all the frames to a one video using imageio
-     path = '/content/video_results/restored_imgs'
-     image_folder = os.fsencode(path)
-     print(image_folder)
-     filenames = []
-    
-     return  "/content/audio.wav"
+     
        
    elif gender == 'Male' or gender == 'male':
       print(gender)