Try a single input component

app.py

@@ -102,11 +102,10 @@ def process_func(x: np.ndarray, sampling_rate: int) -> dict:
 
 
 @spaces.GPU
-def recognize(input_microphone, input_file):
-    if input_microphone:
-        sampling_rate, signal = input_microphone
-        signal = signal.astype(np.float32, order="C") / 32768.0
-    elif input_file:
+def recognize(input_file):
+    # sampling_rate, signal = input_microphone
+    # signal = signal.astype(np.float32, order="C") / 32768.0
+    if input_file:
         signal, sampling_rate = audiofile.read(file, duration=duration)
     else:
         raise gr.Error(
@@ -167,21 +166,21 @@ with gr.Blocks() as demo:
     with gr.Tab(label="Input"):
         with gr.Row():
             with gr.Column():
-                input_selection = gr.Radio(
-                    ["microphone", "file"],
-                    value="file",
-                    label="How would you like to upload your audio?",
-                )
+                # input_selection = gr.Radio(
+                #     ["microphone", "file"],
+                #     value="file",
+                #     label="How would you like to upload your audio?",
+                # )
                 input_file = gr.Audio(
-                    sources="upload",
+                    sources=["upload", "microphone"],
                     type="filepath",
                     label="Audio file",
                 )
-                input_microphone = gr.Audio(
-                    sources="microphone",
-                    type="filepath",
-                    label="Microphone",
-                )
+                # input_microphone = gr.Audio(
+                #     sources="microphone",
+                #     type="filepath",
+                #     label="Microphone",
+                # )
 
                 # output_selector = gr.Dropdown(
                 #     choices=["age", "gender"],
@@ -201,14 +200,13 @@ with gr.Blocks() as demo:
 
             # output_selector.input(update_output, output_selector, output)
 
-        inputs = [input_microphone, input_file]
         outputs = [output_age, output_gender]
 
-        input_selection.change(toggle_input, input_selection, inputs)
-        input_microphone.change(lambda x: x, input_microphone, outputs)
-        input_file.change(lambda x: x, input_file, outputs)
+        # input_selection.change(toggle_input, input_selection, inputs)
+        # input_microphone.change(lambda x: x, input_microphone, outputs)
+        # input_file.change(lambda x: x, input_file, outputs)
 
-        submit_btn.click(recognize, inputs, outputs)
+        submit_btn.click(recognize, input_file, outputs)
 
 
 demo.launch(debug=True)