After executing the [stream_audio.py](https://github.com/vineeths96/Spoken-Keyword-Spotting/blob/master/src/stream_audio.py) with the microphone, I got a result similar to the image below.  Does anyone know why this program needs to be handled like the example below? ``` # Wave plt.plot(data[-len(data) // 2 :]) # Filterbank energies plt.imshow(fbank[-fbank.shape[0] // 2 :, :].T, aspect="auto") ```