Help With Deep Learning Speech Command Recognition

Hello!

I am new to deep learning, and am working on a speech recognition program for my engineering project. I have read through and run the Deep Learning Speech Recognition Example, but I need to train a neural network on several of my own speech commands to connect to an iRobot Create 2. I have two questions about this: firstly, how do I go about training the neural network on my own speech commands? I have many audio files of each command ready for training, but since the commands are different from those in the Deep Learning Speech Recognition Example, and the Example uses commands like splitData which refer only to the example command audio files, I am not sure the best way to go about doing this. Seccondly, once I have the neural net to where it can accurately recognize the speech commands, I want MATLAB to execute a task upon detecting a particular command. For example, one of my commands is "plot data." I will define a set of x and y values beforehand, and I want MATLAB to create a plot of this data when it detects a speaker saying the command "plot data." What is the best way to do this? I apologize as I am still very new to all of these concepts! They are fascinating, but I definitely still have a lot left to learn!

Thank you!