December 09, 2008, 8:55 PM — Cisco will add real-time translation to its TelePresence high-definition conference technology next year, enabling people in several different countries to meet virtually and each hear the other participants' comments in their own languages.
The feature is expected to go on sale in the second half of 2009 with an initial set of 20 languages, said Marthin De Beer, senior vice president and general manager of Cisco's Emerging Technologies Group. It will include speech recognition in the speaker's native language, a translation engine, and text-to-speech technology to deliver the words in a synthesized voice on the other end. Users will also be able to display subtitles if they choose, he said. Both Asian and Western languages will be represented in the initial set, which will later be expanded.
Speaking at a press luncheon at Cisco's C-Scape analyst conference in San Jose, California, De Beer also said the company will deliver a TelePresence system for consumers within 12 months. Trials are set for the middle of next year. A possible application of the home system will be medical visits from a remote doctor. The company's goal is to price that system competitively against large-screen high-definition TVs, he said. Cisco's smallest TelePresence product, the 37-inch System 500, was introduced in May with a list price of US$33,900.
TelePresence is a major focus for Cisco as well as one of its fastest-growing products, according to company executives. Launched in late 2006, it now has at least 250 customers, De Beer said. In public suites and through exchanges offered by carriers, including AT&T, the technology is branching out beyond individual enterprises. One TelePresence buyer is a European airline that wants to set up public meeting suites in airports, according to De Beer. The airline will offer TelePresence meetings as an alternative to time-consuming short-haul flights, he said.
Cisco showed off the future translation capability in a demonstration at C-Scape. Chairman and CEO John Chambers met with a Cisco employee in Spain who spoke in Spanish and had his words translated into English. There was a delay of several seconds before each comment came across as synthesized speech. By the time the product ships, that delay will be reduced to milliseconds, De Beer said.
The feature will be delivered as software for Cisco's Media Experience Engine, announced Monday, which can also adapt video to many different devices and network connections for broader distribution. The translation function will take advantage of a future, chassis-based version of the Media Experience Engine equipped with specialized processors, De Beer said. He said the speech-to-text portion is relatively easy and that the translation software is nearly finished. It will be able to take context into consideration to correctly represent the speaker's meaning.
The translation system will be able to scale beyond two-way meetings, according to Cisco. De Beer gave the example of a TelePresence session among participants from four countries, speaking in four different languages. Each would be able to hear and read everyone's words in his or her own language, he said.
Around the same time that the translation system becomes available, Cisco will offer a transcription feature for the Media Experience Engine to perform speech-to-text conversion of video content. This will create a searchable word-for-word transcript and allow users to go directly to the spot in a video they want to see, the company said.
