Update app.py

app.py

@@ -68,16 +68,16 @@ def inference(audio):
   # Save
   soundfile.write(x_save_path, wav_x, model_sr, 'PCM_16')
   soundfile.write(xrec_save_path, wav_xrec, model_sr, 'PCM_16')
-  return 'vocoded_orig_spec.wav', f'specvqgan_{bitrate:.2f}kbps.wav'
-  
-title = "Anime2Sketch"
-description = "demo for Anime2Sketch. To use it, simply upload your image, or click one of the examples to load them. Read more at the links below."
-article = "<p style='text-align: center'><a href='https://arxiv.org/abs/2104.05703'>Adversarial Open Domain Adaption for Sketch-to-Photo Synthesis</a> | <a href='https://github.com/Mukosame/Anime2Sketch'>Github Repo</a></p>"
+  return 'vocoded_orig_spec.wav', f'specvqgan_{bitrate:.2f}kbps.wav', tensor_to_plt(x, flip_dims=(2,)), tensor_to_plt(xrec, flip_dims=(2,))
+
+title = "SpecVQGAN Neural Audio Codec"
+description = "Gradio demo for Spectrogram VQGAN as a Neural Audio Codec. To use it, simply add your audio, or click one of the examples to load them. Read more at the links below."
+article = "<p style='text-align: center'><a href='https://arxiv.org/abs/2110.08791'>Taming Visually Guided Sound Generation</a> | <a href='https://github.com/v-iashin/SpecVQGAN'>Github Repo</a></p>"
 
 gr.Interface(
     inference, 
     gr.inputs.Audio(type="file", label="Input Audio"), 
-    [gr.outputs.Audio(type="file", label="Original audio"),gr.outputs.Audio(type="file", label="Reconstructed audio")],
+    [gr.outputs.Audio(type="file", label="Original audio"),gr.outputs.Audio(type="file", label="Reconstructed audio"),"plot","plot"],
     title=title,
     description=description,
     article=article,