Talking Computers Just Around the Corner

Thanks to recent innovations, keyboards may soon be taken over by speech-recognition technology as the primary human-computer interface. "Computers are getting smarter, and that is good for speech recognition, but the better news is that society is getting trained," says Serotek CEO Mike Calvo.

Marketability of early speech-recognition products was limited because it required users to modify their speech patterns. Since then, companies have had them tailor-made for specific sectors and develop applications that fulfill particular requirements.

Experts foresee an increased rollout of speech-recognition-equipped mobile devices, considering the handset crowd, which already use the devices for cell-phone calls or dictation and desire better features that could perhaps do away with their reliance on small keyboards. Speech technology capable of identifying emotions is bound to attract considerable attention. Much of the industry is pursuing interactive voice response systems smart enough to use natural-language processing to handle customer calls and vastly improve their accuracy and sensitivity.

Tim Kraskey from Spanlink Communications feels that such systems would not just comprehend the meaning in customers’ statements but also use it to direct the caller or agent to pertinent information. The mainstream adoption of speech-recognition products faces daunting challenges like hesitation among consumers, scalability and throughput issues, effort into customization and the tendency for accents to trip up the technology.

Source: Sci-Tech Today