Tag Archives: Speech Technology

Automated Speech Recognition

edited 16030457_sOne of the most annoying phenomena on our planet is the automated phone system. Everyone knows the technological turbulence inherent in speech recognition technology. Try navigating voice user interface with a speech issue, where the error rate increases with frustration. CLARITY?…puh-leeze!

Here’s the deal: speech recognition software operates on units called phonemes, the basic audibles of a spoken language…only I tug and gnaw at words that seamlessly flow off most tongues, concocting my own unique sounds that become hopelessly lost in mathematical modeling. Communication deteriorates into a comedic exercise of “stabs in the dark,” muse confounding machine. I never understand how my bank’s system confuses “Agent” with “Balance” though they do share the same number of syllables. To be honest, the person who programmed the system probably wouldn’t understand me let alone a computer-generated interface trying to reconcile its algorithms with my zany sound waves.

Pure speech recognition is the ultimate nightmare as I repeatedly press “0” in stubborn defiance of the lack of a human option. Worse, it deprives me of the opportunity to escalate the call to a “Supervisor” lounging in a remote call center. By this point, I require one seriously advanced piece of intelligence – artificial or real – to decipher my garbled rantings.

As for hands-free computing, my arms may desperately yearn for a rest but I envision Sir iMac erasing the hard drive when I command him to open Microsoft Word!