Speech recognition for the Sugar Learning Platform.
##What is this about? Sugar Listens is a GSoC 2014 project that seeks to provide speech recognition capabilities to Sugar Activity developers. For more information, please refer to the project proposal.
##Setup and Run
- Install the dependencies. If you are using Fedora 20 like I am it is as simple as running:
sudo yum install pocketsphinx pocketsphinx-libs pocketsphinx-plugin pocketsphinx-devel pocketsphinx-python pocketsphinx-models git python-setuptools python-lockfile- Clone this repository:
git clone https://github.com/rparrapy/sugarlistens.git- Install the module with setuptools:
cd sugarlistens
python setup.py develop- Sugar Listens uses D-Bus' system bus to communicate with Sugar Activities. For this to work, you must enable it by adding a few lines to the existing policy element of /etc/dbus-1/system.conf:
<policy context="default">
...
<!-- This lines are needed for Sugar Listens to work -->
<allow own="org.sugarlabs.listens.recognizer"/>
<allow send_destination="org.sugarlabs.listens.recognizer"/>
<allow receive_sender="org.sugarlabs.listens.recognizer"/>
<policy context="default">- Run Sugar Listens in a terminal window:
python sugarlistens/sugarlistens/recognizer.py- Test it by running one of the following projects:
- Livedemo: it doesn't get much simpler than this.
- Maze: ready to play a game?
- Sugar: try running your favorite Sugar Activities with your voice. Pretty cool, huh?
##Thanks Current implementation is based on the Pocketsphinx speech recognition engine, so thanks to the guys at CMU that made this possible.