Now that speech recognition is hot on PCs and Smartphones, Ros is at the forefront of technology and certainly not missing out on such a handsome technology. Ros uses code from the CMU Sphinx and Festival Open source projects, issuing independent speech recognition packs, and converting the identified voice to text, and then allowing the robot to intelligently process and speak.
First, speech recognition package
1, installation
Installation is simple, you can use the Ubuntu command directly, install the dependent library first:
$
sudo apt-get install gstreamer0.10-pocketsphinx
$ sudo apt-get install Ros-fuerte-audio-common
$ sudo Apt-get Install Libasound2
Then to install the Ros pack:
$ svn Checkout
http://albany-ros-pkg.googlecode.com/svn/trunk/rharmony
$ rosmake Pocketsphinx
One of the core files is the nodes folder under the recognizer.py file. This file collects voice information through the microphone, and then calls the speech recognition library for recognition to generate text messages, which can be subscribed to by the other nodes, and then processed by the/recognizer/output message.
2, testing
Once the installation is complete, we can run the test.
First, insert your microphone device and test the microphone for voice input in the system settings.
Then, run the test program in the package:
$ roslaunch Pocketsphinx Robocup.launch
At this point, you will see a large segment of information in the terminal. Try to say some simple statements, of course, must be English, for example: Bring Me the glass,come with me to see if it can be identified.
"Ros by Example" This book has been written in the test is very accurate, but I am in the test feel very inaccurate recognition, may be my English is too bad.