Office XP Speaks Out: Voice Recognition Assists Users

REDMOND, Wash., April 18, 2001 -- "I’m absolutely floored by this technology says Robert Hensing, a Microsoft product support specialist who suffers from repetitive stress injury (RSI). Hensing’s job requires him to send and receive daily over 300 e-mails that address complex subjects and require extensive answers. As result of having to type so frequently at work for the past 2.5 years, the  pain in the 25-year-old’s fingers became so bad he couldn’t bear to open a jar of baby food for his infant son or simply turn the ignition key in his car. 

According to the Association of Repetitive Motion Syndrome, (RSI), a painful and debilitating problem that results from performing the same physical motion over and over, may, over time, cause temporary or permanent damage to cartilage, tendons, ligaments, nerves and muscles involved in producing the motion. "I was terrified that I’d done irreparable damage, and I was extremely depressed and frustrated," Hensing says. "After all, I’m only 25 years old and don’t know what I’d do for a living if I couldn’t use my hands to type." Microsoft Office XP, the new version of Microsoft’s desktop productivity suite, can help knowledge workers like Hensing.

The product includes advanced speech-recognition functionality in all Office programs, enabling people to enter and edit data, control menus and execute commands by speaking into a microphone. It will enable customers who suffer from repetitive stress injury and other related injuries to improve productivity by speaking to their PC. Hensing installed Office XP, a sound card and a headset, and after training on the voice-recognition system, began talking instead of typing when using Office applications. Now, Hensing can send and receive e-mail using his voice with minimal keyboard interaction, and also verbally composes large Microsoft Word documents and Microsoft PowerPoint presentations.  "I’m pleased to report that after several weeks, the pain in my fingers has nearly subsided," he says. "Office XP exceeded my expectations, and I can be pretty tough to please." 

Hensing’s story is just one example of how Office XP provides solutions for people with unique usage scenarios, such as those with physical challenges or those who need to communicate by phonetically converting words to characters in a desired language. From people with repetitive-strain injuries to East Asians, for whom typing in their native language has been a laborious process, Office XP’s voice-recognition is transforming the way people work.  Office XP includes innovative technologies developed by Microsoft’s research division to provide speech recognition in three languages: English, simplified Chinese and Japanese. Third parties can also develop Office XP-based solutions for all languages.  Office XP offers two modes of operation: dictation mode, which allows users to dictate memos, letters and e-mail messages; and voice command mode, which allows users to access menus and commands using voice input. 

Speaking to Office XP

"Dictation will help users of U.S. English who are poor typists or who are unable or unwilling to type," says Jeff Reynar, a Microsoft program manager.  "Voice-command mode will make keyboard users more productive by letting them keep their hands on the keyboard while they use Office applications."  While not targeted for skilled typists who prefer the keyboard or for people with certain disabilities, those who benefit from dictation capabilities include poor typists, individuals who must limit their typing due to conditions such as RSI and those who use an Input Method Editor (IME).  For example, people who enter text in languages such as Chinese and Japanese must type out words phonetically and use an IME to convert the words to characters in the desired language 'an inherently slow process'. Written Chinese contains more than 6,000 characters, and for computer users who employ that language, the large character set is an imposing barrier that reduces productivity.  Voice input makes the process twice as fast. Xuedong (X.D.) Huang, general manager of Speech at Microsoft, depends on voice recognition when sending e-mail to friends in China.  "For people in China and Japan, composing a piece of e-mail is like being forced to use a telephone keypad to try and locate thousands of characters," Huang says. "That’s a very inefficient process even with 26 letters. So when you think about the way millions of people in the U.S. and 1.2 billion people in China interact with the computer, they are all handicapped because the keyboard is not designed for their unique language needs."  Huang now uses Office XP speech recognition capabilities built into Outlook to communicate with friends and family located in China, his native country.   "I talk to the computer and I find that the technology is about two to three times faster than the input method editor I used before," he says. "I love this technology."

Office XP - Your Wish Is My Command 

Voice command mode in Office XP supplements traditional mouse-aided command execution with audible instructions that allow people to select menu, toolbar and dialog box items. For example, with a simple voice command, people can:   Add formatting to text by issuing the following vocal instruction:   [Bold.  Center.  Font. Tahoma.]     Open and close various menus and access dialog box functions: [Tools.   Word  Count. Close.]   Issue commands: [File. Print. Current Page. OK.]   "Voice command capabilities reduce dependency on the mouse so that users can  enjoy more uninterrupted focus on the keyboard for data entry," Reynar says.  "This feature is particularly appropriate for experienced typists and   data-entry  personnel whose productivity will be enhanced by keeping their hands on the  keyboard. It eliminates the many transitions they make from the keyboard to  the  mouse to access various menus and commands."

 Office XP makes it easy to switch between the two modes, and the user  interface makes it clear which mode the user is in. When users are in voice command  mode,  however, they cannot dictate text, and when they switch to dictation mode,  only  a few commands can be issued. The implementation of Office XP’s speech  functionality and the two modes is based on large amounts of usability  research,  which found that two modes were usually better than one.   

 "We found that most people that tried to dictate and issue commands at the same time would likely end up with inaccurate results," Reynar says. "For example, they would intend to bold the formatting of a word and instead, the software we were testing would literally type out ‘Bold.’ We decided to take the bi-modal approach to ensure that each mode would work well and so that people could rely on the speech recognition function they were performing."

 Reynar also explains that the right equipment is necessary for the best Speech recognition experience. "We recommend that customers use a high-quality close-talk microphone and audio output device when using the speech recognition functionality," Reynar says. "A Plantronics communication headset is Microsoft’s recommended headset because of its noise cancellation technology, sound quality, comfort benefits and voice recognition accuracy."  

In addition to a close-talk microphone and audio-output device, the speech-recognition functionality in Office XP requires a Pentium II 400MHz or  higher processor and 128 or more MB of RAM.