Hello. I need to incorprorate speech recognition and text to speech in a
robotics application. I am confused as to what I should pursue, Sphinx2, 3, 4,
Pocketsphinx with base, etc. as the docs suggest selecting "the right one".
I forsee background noise control an issue and the recognition of multiple
voices as necessary. The commands will be short utterances, for instance,
"stand up, move forward ten feet, run diagnostics, etc." The plan is to use
openSUSE Linux as a platform and Python as a scripting and general control
program language.
The docs are vague as to how all these packages are related. I wish to avoid
spending months with the wrong software since the overall control program and
mechanical control will be quite involved, not to mention the need to rewrite
any part of the progam due to a poor choice of direciton here.
In summary, the project is in the planning stages so I am simply asking for
pointers.
Thank you.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I need to incorprorate speech recognition and text to speech in a robotics
application. I am confused as to what I should pursue, Sphinx2, 3, 4,
Pocketsphinx with base, etc. as the docs suggest selecting "the right one
You need to use pocketsphinx. You can learn more about CMUSphinx decoders on
our website:
Thank you for responding. I missed the versions page and that answers an
important question.
In "Building the application", during installation, besides an arithmetic
option, we have, "You can also configure with or without python."
I do not know what this means. It is especially important to me since I am
tending towards Python for the overall system language since the code is more
easily read and has the ability to wrap C++ if necessary. So in terms of
Pocketsphinx, if my program language is Python, should I, or do I need to,
configure with Python?
I have practically no experience in programming and no time to waste, so the
best direction to take here would be very helpful and should end this thread.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
In "Building the application", during installation, besides an arithmetic
option, we have, "You can also configure with or without python."
Pocketsphinx has Python support included. It will be built only if python
development headers are installed in your system. You can check if Python was
found in the output of configure. If you don't have Python headers installed
you probably want to install them. You can find some Python examples in
pocketsphinx/python/ps_test.py
I have practically no experience in programming and no time to waste,
Sorry, we aren't that level plug-and-play. We hope to be there one day though.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Ok. This is good. Yes, I understand you are not plug and play. Perhaps one day
I will send you one, with a GUI, too.
I will comment on why I was confused perusing the site. There is really
nothing wrong with it. It simply is not written for a confused, rank beginner.
After perusing some forum threads, I see many of the same, basic questions are
asked and answered over and over again--albeit in different contexts. This
goes on everywhere, at every computing forum. And, as the case was here, the
information is often a page away, or somewhere in a tutorial.
I would like to see a basic guide for the rank beginner. In fact, I may write
one since the process of doing so will reinforce my knowledge and memory,
besides being good for formal notes that may be useful for future reference.
My tutorial would be specific: "Pocketsphinx with Python in a Robotics
Application". If something major changes, such as the programming language or
the incorproation of additional software, I would write another one using the
same format, rather than adding many notes and links.
Again thank you very much. Now I know where to go next.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hello. I need to incorprorate speech recognition and text to speech in a
robotics application. I am confused as to what I should pursue, Sphinx2, 3, 4,
Pocketsphinx with base, etc. as the docs suggest selecting "the right one".
I forsee background noise control an issue and the recognition of multiple
voices as necessary. The commands will be short utterances, for instance,
"stand up, move forward ten feet, run diagnostics, etc." The plan is to use
openSUSE Linux as a platform and Python as a scripting and general control
program language.
The docs are vague as to how all these packages are related. I wish to avoid
spending months with the wrong software since the overall control program and
mechanical control will be quite involved, not to mention the need to rewrite
any part of the progam due to a poor choice of direciton here.
In summary, the project is in the planning stages so I am simply asking for
pointers.
Thank you.
You need to use pocketsphinx. You can learn more about CMUSphinx decoders on
our website:
http://cmusphinx.sourceforge.net/wiki/versions
I think that page is quite clear about which decoder to select. Maybe you
didn't read it beforehand. If you did please explain why you were confused.
I suggest you to check the tutorial which describes some basics of CMUSphinx
http://cmusphinx.sourceforge.net/wiki/tutorial
Thank you for responding. I missed the versions page and that answers an
important question.
In "Building the application", during installation, besides an arithmetic
option, we have, "You can also configure with or without python."
I do not know what this means. It is especially important to me since I am
tending towards Python for the overall system language since the code is more
easily read and has the ability to wrap C++ if necessary. So in terms of
Pocketsphinx, if my program language is Python, should I, or do I need to,
configure with Python?
I have practically no experience in programming and no time to waste, so the
best direction to take here would be very helpful and should end this thread.
Pocketsphinx has Python support included. It will be built only if python
development headers are installed in your system. You can check if Python was
found in the output of configure. If you don't have Python headers installed
you probably want to install them. You can find some Python examples in
pocketsphinx/python/ps_test.py
Sorry, we aren't that level plug-and-play. We hope to be there one day though.
Ok. This is good. Yes, I understand you are not plug and play. Perhaps one day
I will send you one, with a GUI, too.
I will comment on why I was confused perusing the site. There is really
nothing wrong with it. It simply is not written for a confused, rank beginner.
After perusing some forum threads, I see many of the same, basic questions are
asked and answered over and over again--albeit in different contexts. This
goes on everywhere, at every computing forum. And, as the case was here, the
information is often a page away, or somewhere in a tutorial.
I would like to see a basic guide for the rank beginner. In fact, I may write
one since the process of doing so will reinforce my knowledge and memory,
besides being good for formal notes that may be useful for future reference.
My tutorial would be specific: "Pocketsphinx with Python in a Robotics
Application". If something major changes, such as the programming language or
the incorproation of additional software, I would write another one using the
same format, rather than adding many notes and links.
Again thank you very much. Now I know where to go next.