Thank you so much!
I got it!

@cvp Thank you so much!

@JonB Thank you so much! I’m gonna try!

@johnridesabike Based on your sample, I was able to implement my requirement! Thank you so much.

@cvp Thank you so much for your comment. I did’nt know this function. However, as far as I read a description about recognize, recognize function can translate a voice to text from recorded file.
I hope that my application can receive a voice directly, but if I implement following procedure, my hope may be able to be implemented.

Sound recorder is executed. Record a user voice and save it. Call speech.recognize Translate voice file made by No.2 section to text file.

@JonB I finally have done sample program which I want to do. In my recognition, all view is regarded as a component like button, slider on the main canvas. And each components should be added to main canvas with add_subview().
Thank to your comment, I did it.