The Dragon App for iPhone/iPad, Siri, Android’s Google, Dragon for PC & Mac: What Is the Difference?

SeS Icon Speech Voice Recognition There has been increased attention towards speech recognition and voice recognition (and yes, there is a difference) resulting from the recent popularity of Siri, Android’s Google and the Dragon App for iDevices. Those who have never previously considered using speech recognition on their PCs (i.e. Dragon NaturallySpeaking and Dragon Medical Practice Edition) are now curious as a result of the apps on their mobile devices.

However, there is very little information out there to explain exactly what the differences are between all of these interfaces. Every day I field phone calls from clients who are confused by what is possible, and I hope I may be able to explain the differences with this brief post.

Speech Recognition Versus Voice Recognition

In a nutshell, speech recognition is speaker dependent, and voice recognition is speaker independent.

Speech recognition, such as Dragon NaturallySpeaking, Dragon Medical Practice Edition etc. requires you to train Dragon (even if for a couple of paragraphs) how you speak.

Voice recognition, such as the iDevice’s Dragon app, Siri, Android’s Google, voice-activated phone trees, etc. does not require you to train the application and is based on generalized speech patterns.

While voice recognition can be reasonably accurate, especially when the vocabulary available is limited, speech recognition will always be more flexible and often more accurate as a result. I have no doubts that one day the accuracy will be comparable, but that day is far in the future.

Accuracy Variances

While all of the above mentioned products are based on virtually the same programming core, a smaller vocabulary will yield higher accuracy. However, the ability to dynamically train the application will also increase the accuracy.

When using voice recognition applications on your mobile devices (Apple and Android), you will have a limited vocabulary that cannot be modified. This does not make these products ideal for medical, legal or other fields using specialized jargon. Speech recognition applications for your computer (such as Dragon NaturallySpeaking Professional, Dragon Medical Practice Edition and even Dragon Dictate for Mac) allow you to modify the vocabulary as you see fit, thereby adapting to your specific needs.

Speech recognition applications for your computer only require you to train the application for about 5 minutes in order to be extremely accurate, however, we all have vocal variances from the average manner of speaking that require us to sometimes train specific words. Such customization is not available in voice recognition applications for your mobile Devices.

Privacy Concerns

Speech recognition applications for your computer processes your speech locally on the computer on which the application is installed. Voice recognition applications for your mobile devices, on the other hand, transmit your voice to another location over the Internet (i.e. usually a server in the United States), perform the voice to text on another server and transmit the results back to you.

For Canadian physicians and lawyers concerned with privacy, using voice recognition applications would therefore not be prudent. Not only is your voice sent out of the country, your voice (and subsequently, your text) is analyzed and used by Dragon, Google and Apple for quality assurance and future improvement of the technology.

There is a fantastic article entitled “New iPad feature ‘Dictation’ sends/stores private data to Apple servers” by Stephen Chapman that I encourage you to read if you have privacy concerns regarding the use of dictation on mobile devices.  This same information can be applied to Android’s Google voice recognition.

In Conclusion

Voice recognition apps for mobile devices are fabulous helpers for personal use, even limited as they are. However, for medical and legal professional use, speech recognition conducted on your local computer is the most flexible, accurate and also private.

If you need to dictate on the go, consider a digital dictation device such as the Philips DPM 9600 or the Olympus DS5000. These digital recordings can be sent securely from remote locations to your office, where they can be transcribed locally by Dragon.

About Speakeasy Solutions

Based in Vancouver, BC, Speakeasy Solutions Inc. provides Dragon software/licenses, along with installation, support and training services across Canada, from coast to coast.  We carry a variety of microphones for use with Dragon.

We also provide Philips and Olympus digital dictation recorders & transcribers, as well as related services such as installation & setup, and support.

We are in our 12th year of assisting physicians, lawyers and professionals expertly utilize speech recognition and digital dictation solutions towards streamlining office efficiencies.

Contact us any time for further information, a telephone consult or place an order via our website.  We look forward to working with you.

This entry was posted in Speech Recognition, Digital Dictation, Medical, Legal, Dragon, Presales Dragon & Mics, Presales Digital Dictation. Bookmark the permalink.

2 Responses to The Dragon App for iPhone/iPad, Siri, Android’s Google, Dragon for PC & Mac: What Is the Difference?

  1. Is there any speech recognition software for cellphones?

    As you said, voice recognition is so far less efficient than the speech recognition, and I’m always a bit sad to see the gap in accuracy between Dragon NaturallySpeaking on computers and Android’s Google (I don’t use i* devices).