New technologies that enable cell phones to translate speech on the fly and read documents for the visually impaired could have important implications for both educators and students.

Late last year, NEC Corp. announced the development of an automatic Japanese-to-English speech translation tool for mobile phones sold in Japan. The software is aimed at Japanese travelers abroad, but versions for other languages could one day prove useful for educators and administrators in schools with large populations of English-language learners.

And this month, software developer Ray Kurzweil and the National Federation of the Blind (NFB) will begin selling what they say is the first cell phone to incorporate text-to-speech capability.

These developments reveal how quickly text-to-speech and speech translation software is evolving, and they point to a day in the not-so-distant future when students with reading disabilities or language barriers could hold a highly portable solution to these challenges in the palm of their hand.

NFB spokesman Chris Danielsen recently demonstrated the next generation of computerized aids for the visually impaired to reporters.

He fidgeted with his cell phone, holding it over a $20 bill. “Detecting orientation, processing U.S. currency image,” the phone said in a flat monotone before Danielsen snapped a photo. A few seconds later, the phone said, “Twenty dollars.”

The Nokia cell phone is loaded with software that turns text on photographed documents into speech. Besides telling whether a bill is worth $1, $5, $10, or $20, it also allows users to read anything that is photographed, whether it’s a restaurant menu, a phone book, or a textbook.

“We’ve had reading devices before,” Danielsen said, noting similar software is already available in a larger handheld reader housed in a personal digital assistant. Companies such as Code Factory SL, Dolphin Computer Access Ltd., and Nuance Communications Inc. also provide software that allows the blind to use cell phones and PDAs.

Inexpensive hand-held scanners, such as WizCom Technologies Ltd.’s SuperPen, can scan limited amounts of text, read it aloud, and even translate from other languages.

However, the $2,100 NFB device combines all of those functions into one smart phone, said James Gashel, vice president of business development for K-NFB Reading Technology Inc., which is marketing the phone as a joint venture between the federation and Kurzweil.

“It is the next step, but this is a huge leap,” said Gashel, who is blind. “I’m talking to you on the device I also use to read things. I can put it in my pocket and at the touch of a button, in 20 seconds, [I can] be reading something I need to read in print.”

Kurzweil, who developed the first device that could convert text into audio in the 1970s (as well as the current NFB device), said portability is only the first step. Future versions of the device will recognize faces, identify rooms, and translate text from other languages for the blind and the sighted.

The inventor plans to begin marketing the cell phone this month through K-NFB Reading Technology. The software will cost $1,595, and the cell phone is expected to cost about $500, Kurzweil said.

Dave Doermann, president of College Park, Md.-based Applied Media Analysis, said his company is working on similar software for smart phones that could be used by the military for translation and by the visually impaired.

“We don’t anticipate ours being that expensive, but unfortunately we’re not quite to the release yet,” said Doermann, who is also co-director of the University of Maryland’s Laboratory for Language and Media Processing.

Doermann said the company, which has received funding from the Department of Defense and the National Eye Institute, hopes to have its software ready in the next 12 to 18 months.

Kurzweil’s device uses speech software provided by Nuance, said Chris Strammiello, the director of product management at Nuance, who said the company has also developed a prototype reader that uses the internet to access more powerful, server-side computers.

There are about 10 million blind and visually impaired people in the United States, a number that is expected to double in the next 30 years as baby boomers age.

Kurzweil said those with vision problems are not the only ones expected to benefit from the technology. Dyslexics, for example, are expected to be among the users of the current device because of its ability to highlight each word as it’s read aloud, helping them cope with their disability, which affects the ability to read. The highlighting function also can help struggling readers improve their reading skills, he said.

“What’s new here is that both blind people and kids can do this with a device that fits in their shirt pocket,” Kurzweil said.

NFB President Marc Maurer said the device and its PDA predecessor are a “form of hand-held vision” that will make the world of text “much more readily available” to the visually impaired.

Smart-phone speech translation

NEC’s new speech-translation smart phone initially transcribes spoken Japanese into text that it displays on its screen. The phone then converts that text into English. The service can recognize 50,000 Japanese words, but it’s specially designed for smooth translations of common phrases used by travelers, such as “Can I have a subway route map?”

The process of recognizing a Japanese sentence and displaying it on the screen takes about a second, and another second is required for the English version to be displayed. Because the software is operated by a central processing unit and does not require processing by an external server, it can be used in a location where it is not possible to make a phone call.

Although NEC says it is technically possible to have the phone sound out, or “speak,” the English translations, this feature was not included. However, the company says it will continue developing the system, with an eye toward applying the software for translating from English to other languages.

NEC says its software will help bridge boundaries and support communication between those who speak different languages.

But while real-time translation would be “a powerful classroom language tool,” says Ferdi Serim, director of the Online Internet Institute in New Mexico and a board member of the International Society for Technology in Education, having this capability on a cell phone might not be appropriate for K-12 environments, as many schools require cell phones to be turned off or not brought into class.

Serim pointed to another mobile device, called the Phraselator P2—a handheld device that can translate more than 100,000 pre-defined phrases from English into 40 other languages—as possibly a more appropriate model for K-12 schools.

The Phraselator P2 translates in three ways: You can use the Push-to-Talk button for voice-activated speech input, tap the desired phrase on the full-color touch screen, or use the toggle button to scroll and select the desired phrase. The device has an Intel processor, a built-in noise canceling microphone, and a weatherproof speaker.

So far, this handheld translator developed by Voxtec has been used for military purposes and in hospitals, Serim said—but educators should consider its potential for use in schools, too.

School leaders—especially those in New Mexico, the only state that is guaranteed bilingual by the state constitution—need technology that can “provide a bridge that takes people in both directions, not just a ‘moving walkway’ from other languages to English,” Serim said.

Links:

National Federation of the Blind

K-NFB Reading Technology Inc.

Kurzweil Technologies Inc.

Applied Media Analysis

NEC

The Phraselator P2

Online Internet Institute