record their dictation off-line and then submit an audio file for transcription by the app, Mahoney says. But he says Wifi connections are improving.
However, employers will probably need their cybersecurity experts to assess the risks of dictating long documents via a variety of non-company Wifi networks in the field.
Dragon Anywhere doesn’t require new users to spend time training the app to recognize their particular accents or cadences, Mahoney says. The app adapts to an individual’s voice quickly, and it continues to improve in accuracy over time, he says. Users can customize the app by introducing specialized professional terms or place names. These modifications can be shared with an entire staff using Dragon Professional Group, with its Nuance User Management Center.
Dragon Anywhere does not tackle the supreme challenge for voice-to-text technology: the transcription of conversations between two or more people. Writers and researchers still rely on relatively costly transcription services to get full text versions of their interviews, for example. Mahoney says Dragon Anywhere was not designed to transcribe a multi-voice conversation. If a person other than the Dragon user speaks into the device’s microphone, the quality of the transcription would be degraded, he says.
The workarounds to get a multi-voice transcription are a bit cumbersome. The Dragon Anywhere user could repeat another person’s statements into the device microphone, Mahoney says. Or, the second person’s statements could be recorded on an audio file and processed by one of the desktop Dragon versions. The Dragon desktop user might have to habituate the program to the second person’s voice by making corrections to the first minute or so of the transcript, he says.
The lack of fully accurate multi-voice transcribing leaves a niche still open to audio transcription services such as Berkeley, CA-based TranscribeMe, which uses a combination of voice recognition software and human transcribers.
The developers of Nuance’s new suite of synched mobile and desktop products benefited from the company’s experiences with voice-to-text in a number of settings, which allowed them to explore the potential of both mobile and cloud storage, Mahoney says.
Nuance created a Web-based dictation service customized for medical professionals who work at hospitals with low-end computer terminals. In 2012, Nuance launched a free mobile app, Dragon Dictation, that Mahoney says was the first voice-to-text product for mobile devices. It was limited to short sentences, however, like the voice-based texting functions now built into smartphones, he says. Nuance also provided voice technology for consumer electronics makers.
The Dragon Dictation app is now less used than Siri and other voice-to-text mobile features built into iOS and Android devices, Mahoney says. Nuance, which is focusing on the professional market for document creation, is not trying to recapture the consumer market for quick bursts of text communication.
“Free alternatives for sending a quick note will continue to be used,” Mahoney says.
