preventing no speech errors

app.py

@@ -42,6 +42,9 @@ def estimate_height(gender,vad, filepath, imperial):
     signal = read_audio(filepath)
     if vad:
         signal = get_speech(signal)
+        if len(signal) < 1:
+            return f"No speech detected or signal too short!"
+
 
     embedding = torch.squeeze(classifier.encode_batch(signal), 0)
     if gender == "Detect" or gender is None: