Update app.py

app.py

@@ -23,14 +23,19 @@ def get_pipeline_prediction(pil_image):
     
     processed_image = render_results_in_image(pil_image,
                                             pipeline_output)
-    return [processed_image, narrated_text["audio"]]
+    return [processed_image, narrated_text]
 
 demo = gr.Interface(
   fn=get_pipeline_prediction,
   inputs=gr.Image(label="Input image", 
                   type="pil"),
-  outputs=[gr.Image(label="Output image with predicted instances",
-                   type="pil"), "audio"]
+  # outputs=[gr.Image(label="Output image with predicted instances",
+  #                  type="pil"), "audio"]
+    outputs=[gr.Image(label="Output image with predicted instances",
+                 type="pil"), 
+             gr.Textbox(label="Transcription",
+                  lines=3)]
+    
 )
 
 demo.launch()