Dragon Systems Unveils
Revolutionary
Breakthrough with Continuous Speech
Recognition
Dragon NaturallySpeakingTM
Enables Users
to Talk to Their PC Continuously and Naturally
New York, NY -- April 2, 1997 -- A revolutionary breakthrough in continuous speech recognition technology was announced today at a press conference in the World Trade Center, by worldwide leader Dragon Systems, Inc. The Newton, MA-based company, which launched a number of industry firsts in speech recognition for more than a decade, introduces Dragon NaturallySpeaking, the world's first general purpose, large vocabulary, continuous speech recognition product. Dragon NaturallySpeaking enables users to dictate general text more naturally using continuous speech, i.e., without pauses between words.
"When I first started in speech recognition 25 years ago, there was some doubt that this day would ever come," said Jim Baker, CEO and Co-Founder of Dragon Systems. "Now, Dragon is proud to announce the speech product that everyone has been waiting for. This application is ideal for people who need to get their ideas and words rapidly into text, and want to be freed from the tyranny of the keyboard."
With Dragon NaturallySpeaking, users no longer need to type text into a computer. Users simply talk in a natural manner and pace. The spoken words immediately appear on the screen with every word spelled correctly. The powerful text editing and formatting capabilities allow users to bold, italicize, change the font, etc. by voice commands. Documents can be printed or copied into popular applications, such as Microsoft Word, Corel WordPerfect, Qualcomm Eudora, Lotus Notes, etc.
"An achievement of this magnitude is only possible due to the accomplishments and goodwill of our colleagues," said Jim Baker, President and Co-Founder of Dragon Systems. "We have conquered a series of technical challenges, created practical products and launched these into a global economy. Passion and persistence are the primary fuel in this quest. Today, we have the privilege of declaring victory - a victory for us all."
Dragon Systems is also strongly committed to international markets. "Dragon Systems plans to introduce international versions of Dragon NaturallySpeaking later this year," said Andreas Widner, International Marketing and Sales Director.
Features and Benefits
- True Continuous Speech - Dragon NaturallySpeaking allows users to talk naturally and at a normal pace. Unlike "discrete speech" products, which require users to pause between words, users can now speak continuously. Experienced users can expect to speak at aver 100 words per minute.
- Highly Accurate - Dragon NaturallySpeaking is the most accurate general purpose dictation product ever produced by Dragon Systems. With regular use and training, users can achieve 95% or greater accuracy.
- Customizable 30,000 Word Active Vocabulary - Dragon NaturallySpeaking comes with an active vocabulary fully loaded with 30,000 commonly used words and names covering a wide range of topic areas. Users can customize the active vocabulary with the Vocabulary Builder tool which automatically reads documents for the words a user uses often. Additional words can be retrieved from the 230,000+ word back-up dictionary by typing the first few letters of the word during correction. New words can be added simply by spelling them once.
- 230,000+ Word Back-Up Dictionary - The 230,000+ word back-up dictionary is much more than just a spelling list. It contains speaker-independent acoustic and language usage information, which maximizes accurate recognition.
- Natural Spelling - For the first time in any dictation product, users can spell words the way they normally would instead of using a special alphabet. With Dragon NaturallySpeaking, "a" means "a" and "b" means "b."
- Select-and-Saytm - This simple approach makes editing and formatting as easy as dictating. Users do not need to navigate around a document with special movement commands, which are often required in other speech products. With Dragon NaturallySpeaking, users move to and select a word or phrase simply by saying "select" followed by the word or phrase. They can then replace the text by saying a new word or adjust the format of the text by voice, keyboard, or mouse.
- Delayed Correction and Adaptation - Editing and correcting can be done immediately or later. This allows users to concentrate on their thoughts instead of the PC screen as they dictate.
- Dictation Shorthands - Similar to macros, dictation shorthands allow users to replace spoken text with other text as they speak. This is particularly useful if the user likes to dictate an acronym or abbreviation, but prefers to have the full text appear on the screen. For example, the phrase "Triple A" can automatically expand to become "American Automobile Association."
- Fast and Fun Training - Instead of reading from a word list, Dragon NaturallySpeaking's training asks users to spend 18 entertaining minutes reading from Dr. Arthur C. Clarke's newest novel, 3001: The Final Odyssey, or Dave Barry's popular Dave Barry in Cyberspace.
- Multimedia QuickTour and On-line Help with Multimedia - The totally new multimedia QuickTour overview provides examples of how to dictate text, select text, spell words using natural spelling, select words and capitalize them, as well as copy their text to other applications.
- Audio Set-Up Wizard - This built-in wizard automatically checks the audio input and output compatibility and performance. It also quickly adjusts the audio input, if necessary.
- Menu and Dialog Tracking - Users can activate menus and dialogue boxes by voice.
- Vocabulary Builder - Users can improve the performance of Dragon NaturallySpeaking by customizing the vocabulary for their own tasks. The Vocabulary Builder is a tool that makes Dragon NaturallySpeaking almost as accurate on the user's specific topic as having a custom built vocabulary. the Vocabulary Builder reads the user's documents and builds a custom language model based on the scanned text. The Vocabulary Builder automatically adjusts the set of active words by moving words from the backup dictionary. Finally, the Vocabulary Builder identifies any new words found in the user's text and gives the user an option to add those new words to the vocabulary.
- Modeless Operation - There is no need to switch between command mode and dictate mode. This makes it much more natural for the user, so they can concentrate on their work. The system automatically distinguishes commands from dictated text.
- Pronunciation Guesser - Dragon NaturallySpeaking automatically creates pronunciations for new words, even if those words have never been seen before. The Pronunciation Guesser will also build pronunciations for words if the user does not train them.
Pricing and System Requirements
Dragon NaturallySpeaking will be released in two editions: the Personal Edition and the Professional Edition. The Personal Edition enters beta testing in two weeks and the Professional Edition in the summer. First shipments of the Personal Edition are expected in this quarter and will be priced at $695US. Professional versions aimed at vertical markets, will be priced starting at $995US.
For a limited time, all new U.S. and Canadian purchasers of DragonDictate Classic and Power Editions will be eligible for a free beta copy of Dragon NaturallySpeaking Personal Edition. All beta users will be able to to upgrade to the shipping version of Dragon NaturallySpeaking Personal Edition for a nominal fee. Existing registered DragonDictate customers will receive information on discounted purchases of Dragon NaturallySpeaking.
"In addition to announcing a breakthrough technology, we will offer Dragon NaturallySpeaking at a breakthrough price, making this available to the widest number of people possible," said Jim Baker. Dragon NaturallySpeaking supports Microsoft Windows 95 and Windows NT (version 3.51 and 4.0). It requires a 133 MHz Pentium, 32 MB RAM for Windows 95, and 48 MB RAM for Windows NT.
Dragon NaturallySpeaking supports a broad range of built-in audio industry standard sound cards, including Creative Labs SoundBlaster 16 and compatibles, as well as notebooks with built-in 16-bit audio. Proprietary speech cards are not required for either desktops or portables.
Dragon Systems, Inc., The Natural Speech Company, is a worldwide leader in PC speech recognition. Dragon develops and markets high-performance, cost-effective speech and language technology tools that, in multiple world languages (including American English, British English, French, German, Italian, Spanish and Swedish), enable users to create any text, issue commands, and enter and access data simply by speaking. The company also licenses its technology worldwide through developers, distributors, resellers, OEMs (Original Equipment Manufacturers) and ISVs (Integrated Software Vendors). Dragon Systems ports its technology across multiple operating systems and hardware platforms - from hand-held devices and personal computers to workstations, and is pursuing leading-edge research and development projects for computer and telecommunications applications.
Founded in 1982, Dragon Systems is headquartered in Newton, Massachusetts; its Dragon Systems UK Ltd. subsidiary, which is focused on speech recognition for telephony and high-noise environments, is based in Cheltenham, England; Dragon Systems GmbH is located in Glashutten, Germany in the greater Frankfurt/M. area.
Dragon Systems and DragonDictate are registered trademarks of Dragon Systems, Inc. NaturallySpeaking, the Natural Speech Company and Select-and-Say are trademarks of Dragon Systems. All other trademarks are the property of their respective holders.