0
$\begingroup$

Rosanswers logo

I installed the pocketsphinx package on computer A, and tried to run it in various ways, (1) directly run on computer A, i.e., >> roslaunch pocketsphinx voice_cmd.launch. The recognizer node and voice_cmd_vel node can fully load and function well; plus the recognizer node will pop up a long list of info into the window. (2) from computer B, first ssh into computer A, then run >> roslaunch pocketsphinx voice_cmd.launch, and this also works fine as in (1). (3) from computer B, write a launch file to load voice_cmd.launch on computer A, with machine tag set to A. It becomes weird that the recognizer node hangs in the middle of popping out that long info list, and thus the speech recognition can't continue.

Any tips/suggestions are welcome


Originally posted by clark on ROS Answers with karma: 393 on 2011-09-14

Post score: 0

$\endgroup$

1 Answer 1

0
$\begingroup$

Rosanswers logo

You can look here for potential solutions to your problem.


Originally posted by DimitriProsser with karma: 11163 on 2011-09-14

This answer was ACCEPTED on the original site

Post score: 1


Original comments

Comment by clark on 2011-09-16:
Ok, via rxconsole, the recognizer node is found still running, but the output is not directed to the screen. The problem now evolves into that, whether it is possible to enable node output in local console in case of remote roslaunch... which looks impossible at the moment according to this thread, http://answers.ros.org/question/1688/question-about-roslaunch-and-remote-processes

Comment by clark on 2011-09-14:
unfortunately that should not be the cause. Since in /etc/hosts on both computers, the alias naming have been added for both computer A and B. And for some other packages, the above method (3) works fine except pocketsphinx. So I am wondering what special requirements that this package might have....

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.