“Call Alicia’s mobile number”
“Text message to Bryan & Jason – Congrats on the Giants winning the World Series”
“Find the nearest Starbucks to meet with Kristi”
“Get directions and map from home to Houston’s restaurant near Lenox Square”
“Make a reservation at BJ’s to have dinner with Jennifer Thursday at 7:30 pm.”
“Email to Jill – Love to the grandkids – hug each of them for me.”
“Look up Bruce Grant in Contacts”
These are the kind of things that new mobile voice recognition applications can do or will be able to do in your SmartPhone. Companies like Nuance, Vlingo, Yap and others have developed some sophisticated technology to process spoken commands and convert speech to text in your mobile phone.
This is a difficult problem to solve. It wasn’t too long ago that companies like IBM were using large mainframe computers to process speech. Then, fairly good processing came to the PC. Over the past few years, very good speech recognition systems have been created for mobile devices.
To be sure, many systems use a ‘client/cloud’ model in which the speech is recorded and pre-processed on the phone. Then, it’s sent off to the supplier’s more powerful system that does the ‘heavy lifting’ and the results are sent back to the phone for display and use by the subscriber.
Today, most SmartPhone users can download a voice recognition and processing application that can either do a lot of things (Nuance) or do more specific things like search or navigation. Here are the typical capabilities of today’s voice recognition and processing in mobile phones:
• Speak a reply to a text message that will automatically convert speech to text.
• Find someone in the Contact list and dial that person’s cell phone number.
• Search for something (uses speech analysis plus a search engine).
o What’s the temperature in New York?
o What was the last closing price for Exxon?
o Who won the World Series on 1925?
• Request navigation information (turn-by-turn directions, maps).
• Compose an email and have it sent to a number of people.
While I think that voice processing is very useful and beneficial in mobile devices (as well as adding to safety while driving), it’s important to remember that there are some situations in which voice processing isn’t appropriate. Three examples: 1) with a date at a restaurant, 2) with a group of people at a party and 3) in a meeting at the office. It simply wouldn’t be appropriate to start a voice request using your phone in these situations.
Most voice recognition services start by pressing a button to begin the voice activated application. This allows the application to ‘pay attention’ only when necessary. I recently had a briefing update with Todd Moser of Sensory that has developed a low power logic that can sit in the background and listen for the key activation phrases. This enables voice recognition to be available always and eliminates having to take the time to begin the voice processing application.
Voice processing is now accepted by most SmartPhone users. While it might have been thought ‘odd’ to give commands via voice to your phone a few years ago, most people today realize that there are real convenience benefits in asking the phone for information or to do something. In this way, SmartPhones are going to become virtual assistants that may get very intelligent over the coming years.
If you haven’t tried out a voice activated service on a SmartPhone, I recommend that you do so. You’ll find that it works well most of the time and will definitely save time. Before long, you’ll be telling all your friends how easy to use and productive this capability really is.
Written By:
J. Gerry Purdy, Ph.D.
Principal Analyst
Mobile & Wireless
MobileTrax LLC
gerry.purdy@mobiletrax.com
404-406-5309
Disclosure Statement: From time to time, I may have a direct or indirect equity position in a company that is mentioned in this column. If that situation happens, then I’ll disclose it at that time.