To use the pre-trained audio recognition model in the TensorFlow tutorial (https://www.tensorflow.org/tutorials/audio_recognition) or its retrained model we described before, we'll reuse a listen.py Python script from https://gist.github.com/aallan, and add the GoPiGo API calls to control the robot movement after it recognizes four basic audio commands: "left," "right," "go," and "stop." The other six commands supported by the pre-trained model—"yes," "no," "up," "down," "on," and "off"—don't apply well in our example, and if you want, you can use a retrained model as shown in Chapter 5, Understanding Simple Speech Commands, to support other voice commands for your specific task.
To run the script, first download the pre...