CIO —
Cell phones won’t instantly become simple to use and mobile data services easy to find after the CTIA Wireless trade show in Las Vegas this week, but vendors will demonstrate technologies that are coming this year to solve those problems.
For at least two companies at the show, the answer is voice. Both VoiceSignal Technologies and Nuance Communications will introduce and show off speech recognition systems for SMS (short message service) dictation, search, and other functions, promoting them as easier alternatives to numeric keypads and tiny browsers.
The idea behind these systems is that they can recognize normal speech and not be restricted to a narrow vocabulary. Users can talk at their regular rate and their words will be converted into text, which they can then correct on the phone if necessary, the vendors said.
VoiceSignal’s VoiceMode 2.0 is a big step ahead of the first version of the product, according to Chief Executive Officer Rich Geruson. With the previous product, users had to speak slowly so the system could recognize one word at a time. Now, a user can just pick up the phone, give a voice command to send a text message to a particular contact, and then start talking as they normally would, he said. The technology recognizes components of words, called phonemes, so just from those sounds it can construct words it doesn’t know, he added. The new version also will be expanded from American English to support British English and European languages, the company said.
A cell phone maker will introduce VoiceMode 2.0 on a handset by the third quarter, and that device will probably be introduced first in Europe, according to VoiceSignal.
VoiceSignal also will demonstrate VSearch, a system for making Internet search engines usable via voice commands. It converts spoken search terms into text for entry on a server, then brings back the results in text on the phone’s screen, Geruson said. If the results bring up items with phone numbers, such as businesses, the user can identify one of those results by its number on the list and verbally tell the phone to dial the number of that business.
VSearch could work with a variety of search engines, Geruson said. VoiceSignal is in talks with search engine companies and mobile operators about delivering the system on future cell phones, he said.
Nuance, a longtime maker of speech recognition products for desktops, will kick off a broad strategy for the mobile space called Nuance Mobile. An early product of that plan, to be demonstrated at the show and expected on handsets in the U.K. this year, is Dragon Mobile Dictation. The software supports continuous speech and uses a wide vocabulary, though it is optimized for words people are likely to use in text messages, said Mike Thompson, vice president and general manager of Nuance’s telecommunications and premium services division. Like VoiceSignal’s system, it runs entirely on the handset, he said. However, Nuance later plans to introduce speech recognition systems for other applications, including search and IM (instant messaging). Those will be client/server systems, with some elements of speech recognition taking place over the network, Thompson said.
