Voice to Text Providers

Speech-to-text technology has advanced rapidly in the 21st century, and there are dozens of solutions available. Generally, there are two types of speech-to-text methods: solutions where a person will transcribe the audio manually, and automated speech-recognition systems in which the transcription is done by software.

Benefits

Voice-to-text solutions offer users more productivity because they can dictate notes or essays, and some solutions can translate spoken words into text within seconds.

Automated Solutions

Nuance Communications Inc. offers software applications such as Dragon NaturallySpeaking and MacSpeech Dictate. It also offers an iPhone application, Dragon Dictation, in which the users speaks into the phone for 30 seconds at a time and the words are translated. Another provider, Jott, allows users to call a number and the company's software will transcribe the words into text. Google Voice offers a free voicemail transcription service using its speech-recognition technology.

Human Transcription

Among the providers offering services where humans manually transcribe audio are Purple Shark, Kedrowski Transcription and CastingWords. These services are often used by educational institutions and academic researchers who have large amounts of audio to transcribe.

Considerations

Speech-to-text software solutions are not 100 percent accurate. For instance, a 2007 PC World test of NaturallySpeaking software found that it was between 75 percent and 95 percent accurate, mostly with limited background noise.

References

• PC World: From Speech to Text
• The Chronicle of Higher Education: 5 Easy Speech-to-Text Solutions
• The New York Times: Free Speech (Recognition)