I want to create a system so that when a human being says a word or command through a microphone, such as "shut down", the system can execute that command "shut down".
I used the Deep Speech algorithm on the Persian language database, which takes audio data through a microphone and returns the text. The problem I have now is what I have to do from this point on. I need to explain that my system has to work offline and also the number of commands I have is limited.
The problem you state is a well known problem, and it is called "keyword spotting" os KWS. If you add a wake up word before it (like "hey google/siri"), you can also use "voice command" system to alleviate the problem.
There are two kind of KWS systems: those which develop to detect a hard coded set of keywords, and those who are flexible to amend the keyword set. As you state that you trained an acoustic model using deep speech, it looks that you choose the latter approach.
If you trained a deep speech model, you done the easy half of the solution. The other half which is more tricky is to develop a search algorithm based on your acoustic model. I think wfst based algorithms and prefix beam search algorithms are what you should be looking for.
Regards
