Speech recognition in Linux
There are currently several speech recognition software packages for GNU/Linux, some of them are open-source and others proprietary software.
Native Linux speech recognition
History
In the late 1990s, a GNU/Linux version of ViaVoice (created by IBM) was made available to users for no charge. However, the free SDK was removed by the developer in 2002.
Current development status
Recently, there has been a push to get a high-quality native GNU/Linux speech recognition engine developed. As a result, numerous projects dedicated to creating GNU/Linux speech recognition solutions were established. One major hurdle is the compilation of a speech corpus to enable production of acoustic models. In response, VoxForge, which aims to collect transcribed speech for the use with free and open-source speech recognition engines under the GPL license, was set up.
Ubuntu is currently gathering ideas for implementing speech recognition.[1]
SpeechRecognition Concept
Record an audio stream on your linux machine. Now you have two options:
- process the voice recognition on your local machine or
- submit the audio file to a remote server for converting the audio file into a text string.
The second option (see OpenMoko Speech Recognition) is used mainly on smartphones, because they do not have the performance and disk space to process the speech recognition on the phone.
Free Speech Recognition Engines
The following is a list of current projects dedicated to implementing speech recognition in Linux, as well as major native solution:
- CMU Sphinx is a general term to describe a group of speech recognition systems developed at Carnegie Mellon University.
- Julius is a high-performance, two-pass large vocabulary continuous speech recognition (LVCSR) decoder software for speech-related researchers and developers.
- Simon (This project aims at being extremely flexible to compensate dialects or even speech impairments; requires HTK and Julius)
Related projects:
- Platypus is a open source shim that will allow Dragon NaturallySpeaking running under wine to work with any linux x11 application.
- Vedics is a speech assistant for GNOME Environment
- Xvoice (requires ViaVoice to function)
- GnomeVoiceControl is a dialogue system to control the GNOME Desktop that was developed in the Google Summer of Code in 2007.
- CVoiceControl is a KDE and X Window independent version of its predecessor KVoiceControl
- SphinxKeys lets you essentially type keyboard keys and mouse clicks by speaking into your microphone. It's simple but works pretty much out of the box.
- Open Mind Speech a part of the Open Mind Initiative and aims to develop free (GPL) speech recognition tools and applications, as well as collect speech data.
- PerlBox is a perl based control and speech output.
- VoxForge is a free speech corpus and acoustic model repository for open source speech recognition engines.
It is possible, though complicated, for advanced developers to create Linux speech recognition software by using existing packages derived from open-source projects.
Proprietary Speech Recognition Engines
- Wizzscribe SI: Is a commercial speech recognition server for Linux, launched by Wizzard software in 2006.
- Verbio ASR is a commercial speech recognition server for Linux and windows platforms.
- DynaSpeak from SRI International (speaker-independent speech recognition software development kit that scales from small- to large-scale systems, for use in commercial, consumer, and military applications)
- LumenVox Speech Engine is a commercial library for Linux and Windows for inclusion in other software. It has been integrated into the Asterisk private branch exchange system [2]
Voice control and keyboard shortcuts
Speech recognition usually refers to software that attempts to distinguish thousands of words in a human language. Voice control may refer to software used for sending operational commands to a computer or appliance. Voice control typically requires a much smaller vocabulary and thus is much easier to implement.
Simple software combined with keyboard shortcuts, have the earliest potential for practically accurate voice control in Linux.
Running Windows speech recognition software with Linux
Using a compatibility layer
It is possible to use programs such as Dragon NaturallySpeaking in Linux, by utilizing Wine, though some problems may arise, depending which version is used.[3]
Using virtualized Windows
It is also possible to use Windows speech recognition software under GNU/Linux. Using no-cost virtualization software, it is possible to run Windows and NaturallySpeaking under Linux. VMware Server or VirtualBox support copy and paste to/from a virtual machine, making dictated text easily transferable to/from the virtual machine.
See also
References
External links