Update app.py

app.py

@@ -226,31 +226,25 @@ def predict_category(que, input_image):
     return preds[0]
 
 
-def combine(question, audio, input_image):
-    if question:
-        que = question
-    else:
-        que = transcribe_audio(audio)
+def combine(audio, input_image):
+    que = transcribe_audio(audio)
+    # que = "What is the animal here?"
 
     image = Image.fromarray(input_image).convert('RGB')
     category = predict_category(que, image)
 
     answer = predict_answer(0, que, image)
 
+    # print(category)
+
     tts = gTTS(answer)
     tts.save('answer.mp3')
     return que, answer, 'answer.mp3'
 
 
-model_interface = gr.Interface(fn=combine, 
-                               inputs=[gr.inputs.Textbox(label="Text Question"), 
-                                       gr.inputs.Microphone(label="Audio Question", type="audio", source="microphone"),
-                                       gr.inputs.Image(label="Upload the image")], 
-                               outputs=[gr.outputs.Text(label="Transcribed Question"), 
-                                        gr.outputs.Text(label="Answer"), 
-                                        gr.outputs.Audio(label="Audio Answer")])
-
 
+# Define the Gradio interface for recording audio and displaying the transcription
+model_interface = gr.Interface(fn=combine, inputs=[gr.Microphone(label="Ask your question"),gr.Image(label="Upload the image")], outputs=[gr.Text(label="Transcribed Question"), gr.Text(label="Answer"), gr.Audio(label="Audio Answer")])
 # image_upload_interface = gr.Interface(fn=upload_image, inputs=gr.Image(label="Upload the image"), outputs="text")
 
 # Launch the Gradio interface