Dragon Systems Inc. plans to release the next version of its NaturallySpeaking speech recognition software at the end of this month in American English, with other languages added at the end of the year.
NaturallySpeaking Version 4 will be able to "learn" the voices of users in about five minutes. It also allows users to control Windows desktop applications by voice commands; has new set-up wizards to simplify that process; and enables Web browsing via voice commands, according to the Newton, Massachusetts, company.
Version 4 has an active vocabulary stored in memory of more than 160,000 words or twice the size of the largest active vocabulary now available, according to Dragon Systems, and includes technology that automatically adds to the vocabulary words that users commonly employ. It also enables Web browsing by voice commands, Dragon Systems said in a statement today.
The new version will work on Windows 98, Windows 95 and Windows NT operating systems. It will require a 200MHz Pentium or equivalent processor and at least 48M bytes of RAM for Windows 98 or 95 and 64M bytes of RAM for Windows NT. A CD-ROM and a minimum of 200M bytes of free disk space is required for installation. The software also needs a Creative Labs SoundBlaster 16 compatible Version 4 sound board supporting 16-bit recording.
Dragon's NaturallySpeaking software product line includes editions for teenagers, mobile users and the medical and legal industries. The various editions will be released at different times through the rest of the year, with the first set, including the basic versions with wider applications, due out by Aug. 30 in American English.
The modules in the NaturallySpeaking product line range in price from US$49 for a basic consumer-oriented package and the teen version, to $99 for the standard package, $149 for the business-oriented package, all the way up to $695 for a professional version and $995 for the legal or medical versions.