add audio preprocessing to make sure asr input is single channel audio

app.py

@@ -121,6 +121,9 @@ def transcribe_audio(audio_input):
         audio_array = audio_input[1]
     else:
         audio_array = audio_input
+    # Ensure input is mono by averaging channels if necessary.
+    if audio_array.ndim > 1:
+        audio_array = np.mean(audio_array, axis=-1)
     result = whisper_pipe(audio_array)
     return result["text"]
 
@@ -156,8 +159,7 @@ with gr.Blocks() as demo:
     
     with gr.Row():
         # Audio and Text inputs – only one will be visible based on mode selection.
-        # Changed gr.Audio type from "filepath" to "numpy" to capture audio in memory,
-        # which helps release the recording resource faster.
+        # Use gr.Audio type "numpy" for in-memory capture.
         audio_input = gr.Audio(sources=["microphone"], type="numpy", label="語音輸入 (點擊錄音)")
         text_input = gr.Textbox(lines=2, placeholder="請輸入文字", label="文字輸入")
     
@@ -186,7 +188,7 @@ with gr.Blocks() as demo:
     with gr.Row():
         tts_output = gr.Audio(type="numpy", label="TTS 語音輸出")
     
-    # Button event triggers the classification. Gradio will show a spinner during processing.
+    # Button event triggers the classification.
     classify_btn.click(fn=classify_intent, 
                        inputs=[mode, audio_input, text_input, model_dropdown],
                        outputs=[transcription_output, classification_output, tts_output])