Voice Recognition Learns to Listen to Volcanoes and Geysers' 'Words,' 'Sentences'
February 08, 2012
By David Sims, TMCnet Contributing Editor
It’s always cool when this business technology we write about here gets used for out-of-the-box stuff, like military needs, space technology or going into Yellowstone National Park to help collect seismic data.
According to a recent article in the Billings Gazette, it all happened as just one of those things that just... happens.
Phil Dawson, a geophysicist with the U.S. Geological Survey based out of California, works with the Norris Geyser Basin in Yellowstone. One day a visiting researcher from Spain, Carmen Benitez, happened to need a desk, and it happened that the USGS had a spare one. It happened that this researcher specialized in voice recognition, it happened that Dawson was hanging around, it happened that they started talking, it happened that Benitez decided to see if voice recognition could help sort out data from the Norris Geyser Basin.
As it just so happened, it could. Talk about “natural” speech recognition.
Crediting the recent improvements in the speed of computer processing, Hank Heasler, another geologist working in Yellowstone, said seismologists had been trying to analyze data from the Norris since 2004, but the “new viewpoint” Benitez brought was a key to unlocking the problem.
The trick, Dawson said, was in regarding the hydrothermal movement in geysers, as well as magma movement in volcanoes “as a sentence or a song,” listening to the pop of bubbles in both as the fluids pulse and bump into earth. They’d been collecting the data for about eight or nine years, but couldn’t make sense of it, couldn’t find the “words,” the Gazette wrote, which repeat to indicate an event, until using Benitez’s voice recognition technology.
The idea is to build a “vocabulary” of seismic event “words” that can tell us when something like an eruption or other unusual activity is brewing. There are lots of other applications for analyzing seismic data as well, of course -- most of which probably haven’t even been thought of yet.
It’s another example of how speech recognition technology is being adapted to uses undreamed of not long ago. Last November TMC (News - Alert) noted “the merger of the advantages of cloud computing with state-of-the-art mobile computing technologies that promise to change 21st-century definitions of communication,” and that the technology has other fascinating possibilities in the medical field, among others.
Cool. Okay, now back to your call centers.David Sims is a contributing editor for TMCnet. To read more of David’s articles, please visit his columnist page. He also blogs for TMCnet here.
Edited by Juliana Kenny