Hello,

I have a problem with using simon speech recognition and I hope somebody can help me for my study work.

My problem: I use simon as a speech recognition program for a plane simulator. Every Command the pilot says consists of three parts (Name, command and an additive).
To save the command, we are using the option " Executable Commands". So we produce three text files every command. But the problem is that simon don't understand all of the three parts if we speek the command (the sentence) fluently, so we have to take a short brake between each part of the command.

Is there a possibility to speak a sentence fluently and simon recognize the three parts of the command and save these as a text file, or to decompose the sentence to words?
Or do you have another idea to solve this problem, so that I can speak the command fluently without a break between the words ?

Thanks for your answers.

There has been lots of buzz around tensorflow which I guess might be more advanced now. Have you tried this?

https://www.tensorflow.org/tutorials/se ... ecognition

Thanks for your answer, but we have already built up the whole structure with Simon. What we lack is the decomposition of sentences into single words which in turn trigger their assigned commands. We got it working by speaking each word separately, but that's not an intuitive way.