SpeechLion is a small speech recognition application based on Sphinx-4 which demonstrates desktop command and control.
Currently it is mainly useful for assistance when web-browsing and can save a lot of keyboard and mouse usage. It also has a small Emacs mode and a Unix shell mode. It is developed and tested under Linux and Gnome. The engine runs entirely on the Java VM, and should easily run under Windows. The grammar files (which control most of the features) are dependent on the keyboard shortcuts used by the Mozilla family of browsers under Linux. It is tested with Firefox.
SpeechLion is alpha-quality code. It is basically a proof of concept right now. It does some useful things for me and is at a point where others may enjoy it too.
- Web browsing shortcuts
- Emacs shortcuts
- Shell shortcuts
- Window management shortcuts
- Volume and muting commands
- ssh commands
- Command mode, spelling mode, mousing mode, and off mode
- Low-level keyboard input for most possible keyboard combinations
- Low-level mouse input for motion, button presses, and wheel movement
- Repeat command
- With optional FreeTTS library, provides spoken acknowledgements of commands
SpeechLion requires the following software to run:
- Java runtime environment 1.6 or greater
- Jython 2.2.1 or greater
- Sphinx-4 1.0beta4 binary distro or newer
- Optional: FreeTTS 1.2.2 binary distro or newer
The keyboard shortcuts are tested on a Linux GNOME environment with the Firefox web browser; however, SpeechLion has no dependencies on the GNOME or KDE windowing environments.
Detailed installation instructions are provided in the INSTALL file. Usage instructions are provided in the USAGE file. HTML-formatted versions of all the documents are located in the html directory. Start with html/index.html.
SpeechLion is licensed under the GNU General Public License (GPL). See the COPYING file for details.