Hello, I am trying to implement a device which will recognize gestures, and use these gestures to generate commands for Alexa. It seems that the speech recognition and natural language processing are bundled together into the SpeechRecognizer API. Is it possible to separate the two processes, so that I could still send voice commands, but also occasionally send text (based on gestures) instead and still receive a response from Alexa? Essentially, I want to know if Alexa's usability is limited to voice commands.
I suppose I could instead have a set of pre-generated voice commands for each of the gestures, but I would need to modify the raspberry pi sample app to allow me to use canned audio when a gesture is recognized.