Hayabusa: Composing with Voice Recognition Technology

Hayabusa made his name as a professional wrestler. But his wrestling career was abruptly cut short by an accident in the ring that resulted in severe damage to the cervical vertebra. He was initially paralyzed completely, but, through rehabilitation, gradually regained enough movement to stand up unassisted. Although he still experiences partial paralysis, Hayabusa is now able to use a computer for a range of creative endeavors including writing lyrics, composing music, and illustrating.

Voice-driven Keyboard

Hayabusa's accident occurred during a wrestling match at the Korakuen Hall, the home of combat sports in Japan in October 2001. The damage to his cervical vertebra was so severe that he was at one point on the brink of death, but after 525 days in the hospital and with his commitment to ongoing rehabilitation he eventually regained his ability to stand up without assistance. Still, Hayabusa experiences paralysis that requires assistive technology to use the computer.

For text input, Hayabusa uses Windows Speech Recognition included in Windows Vista. "I can use the mouse for clicking, so most of the basic operations can be done that way," he notes. "But since I can't operate the keyboard, I use the speech recognition feature to input long passages of text." Although Windows can display a software keyboard on the screen, Hayabusa finds pressing individual keys to be slow and unwieldly, so he prefers speech recognition.

Hayabusa used other speech recognition packages before purchasing Windows Vista, but was pleased to find this feature built-into the operating system. "For me it was significant that speech recognition was part of the purchase price. As well as saving me money, it meant that I didn't have to go through all the installation and initial setup procedures. And Windows Vista offers a wide range of options right out of the box."

For voice input, Hayabusa uses a microphone mounted on a stand. He admits that a headset would be a better solution, but says he can't put it on because of difficulties in raising his arms to his head. Since he is unable to get close to the microphone, ambient noise can be a problem, in spite of his best efforts to keep surrounding noise to a minimum. Although recognition errors sometimes occur, Hayabusa says that Windows Vista represents a distinct improvement over previous speech recognition software because it offers a good selection of error correction methods.

For instance, there are several different voice commands that can be used to select text for correction. The user can select the text that has just been entered with the command, "Select the text just entered," then modify it as required. Or, the user can modify the text directly with the command, "Modify" followed by the text you want to modify, which brings up the Modify dialog box. It is then simply a matter of selecting the correct option from the list of suggestions. "The more you use speech recognition, the more it gets used to you," enthuses Hayabusa. "Recognition accuracy goes up and corrections become quicker and easier."

New Efficiency Improvements in Vista

Unique features in Windows Vista are helping Hayabusa to become more efficient at using his computer. For instance, Windows Vista speeds up the process of converting Japanese characters during text input by displaying a list of suggested characters in a dialog box. The new Windows Aero technology in Windows Vista renders the dialog box semi-transparent so that it does not block out the screen. This allows the user to view previous text in the document while selecting a character, which is especially helpful for longer sentences. Another useful feature included in Windows Speech Recognition is the ability to attach numbers to operations on the screen. Again, the numbers are rendered in semi-transparent form so they do not obstruct the controls being displayed beneath them.

Hayabusa, who spends long periods at the computer running multiple programs at the same time, is particularly taken with the new Live Taskbar thumbnails generated by Windows Aero. Placing the cursor over a window button in the Taskbar generates a reduced image of the window, allowing him to view the contents at a glance. For those with limited mobility in the hands, each and every operation that can be combined or omitted in this way constitutes a significant efficiency improvement.

Speech recognition technology helps Hayabusa to achieve his creative potential

"Stage and theatre performance involves speaking directly to the audience," explains Hayabusa. "I produce all my scripts using voice recognition technology, and to me this is a more natural way of getting my ideas across than typing on a keyboard. Because I'm using my voice, just like in an actual performance, I get a better sense of the rhythm and cadence of the words. I think that speech recognition will make it easier for anyone to start writing and composing, and I hope that more people get interested in it."

Limitless Potential

Hayabusa has "fingers in many pies," yet he still relishes a new challenge. One of his chief aims is to write a book that brings together all of his remarkable experiences to-date and his expansive blog entries, using the power of voice recognition technology to create an extended work entirely in his own words. "I want to bring my story to the widest possible audience," says Hayabusa, and it is information technology and particularly the Internet that can help him reach his goal.

Meanwhile, Hayabusa is busily working on the launch of a new fashion label with his own clothing designs, inspired by endless frustrations in trying to locate suitable clothing. Here again, IT plays an integral role. "It's very hard to find stylish and comfortable clothes for people with physical disabilities, and selling them is even more problematic," sighs Hayabusa. He plans to retail his clothing range via the Internet, which will reduce inventory risk and also allow for customization of individual items to order. "It's important to provide choice," he explains. "IT is fantastic in that it broadens the range of choices in so many ways."

As for the future, Hayabusa looks forward to future advances in information technology. "I'd really like a computer that could understand the subtle nuances of human expression and imagery," he confesses. For one who is dedicated to the process of rendering artistic expressions in electronic form, Hayabusa well understands the inherent difficulty of trying to get his ideas down on the computer. "I have difficulty using my hands, so it can be hard to achieve subtle nuances and inflections with the mouse, for instance when I'm doing an illustration. If only the computer could interpret the tone of my voice and render that in a drawing!"