No Jitter is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Translating Telepresence

It seems no matter how spiffy video conferencing and telepresence gets, if you find yourself face to face with someone who doesn't speak your language, that will pretty much put the kibosh on any sort of meaningful interaction you hoped to have. And while video communications may reduce the need for business travel, setting up sessions with participants joining in from different continents means that at least some of them will be joining outside of their regular business hours...sometimes way outside of their regular business hours.As it turns out, a resolution for at least one of these issues is in the works. Researchers at Carnegie Mellon University have apparently paired speech-to-text and real-time foreign language translation software with Cisco TelePresence systems. The result, as it stands now, was demonstrated during John Chambers' keynote address at this year's C-Scape conference in San Jose.

Here's how it worked. John and sidekick Jim Grubb--Cisco's "chief demo officer"--were on stage and sitting in front of a large monitor that I think was part of a Cisco TelePresence 1000 System. On screen was an employee in Madrid who was asked a question I failed to jot down. He responded in Spanish, speaking slowly and carefully, parsing his sentences into short, manageable phrases. Sometimes the solution worked surprisingly well. "Let me think about that a moment," said a synthesized voice, the sort that you'd imagine hearing from Stephen Hawking. At the same time, the text of what the Madrid worker said appeared below his image on the monitor, similar to a subtitle on a foreign film.

At other times, the translation was pretty sketchy. "You can also use to notify local events," was how one sentence came out. This is more like my experience with software-based translations, the sort of awkward, confusing, often nonsensical stuff spit out from online translators like Babelfish. I'd be reluctant to use it in any kind of professional situation, as I have a hard enough time making myself clear in English without wondering how some software is translating me. I'd be thinking, "Are those knit brows because I didn't make my point clearly enough...or is it because my Japanese colleagues just heard me say, 'May I mambo dogface to the banana patch?'"

This is clearly not stuff that's in any way ready for prime time, as folks at Cisco are the first to admit. Apparently there were some serious discussions within the TelePresence team about whether to do the real-time translation demo in the first place. Personally, I'm glad they did. I'm impressed that the technology worked at all, and that both verbal and text-based translations were simultaneously delivered. It's good to see Carnegie Mellon and Cisco push the envelope in terms of what can be done with communications technology as it stands at present, and provide a glimpse of what might be possible a few years from now. Now if they can just do something about those pesky time zones...