speech – the final front ear (or technology for the determined)
Roger Frost tests speech recognition software (199x)
Speech – the final frontier! The efforts to make a computer do as we speak, to type things out for us at least, are making progress. It is good progress too – you can take a good machine, talk boldly to your word processor and no longer will they come to take you away. No longer must you culture the flattest, unenthusiastic voice to use speech recognition software. If you talk in a clear voice, the keyboard can be left alone and the fingers parked away.
But what is a luxury for most folk, fabulous productivity for others may be a salvation for many. More than helping those with a physical disability there is an explosion of promise to those with special needs – from dyslexia to visual impairment. From this ricochet other gains that aid literacy, diction and even thinking.
It is not snake oil, but after talking it up, the expectation must be set: speech recognition works best on the machine you’ll use all the time and better still when you’re proficient. Like learning to touch type, its needs persistence so you will slow down before you speed up. And like any situation with learners, those offered a taste of salvation, will need encouragement to get over the initial blues.
People say things like “speech recognition – yeah it really works” and they’re right. Or sort of. Tell the machine that same claim and it might type, “Each patient care they were” instead. Repeat it several times to the point of frenzy; fart or tap dance – the result’s no worse: sense cannot be had from the thing. Yesterday the machine seemed intelligent, but just now the microphone volume is awry, the sound software is sick, the phone’s ringing and you’re shouting!
The software may come in two types – there are discrete speech recognition packages that require a less powerful computer to process individual words. Many school situations, younger children say, point to this as best. Other packages prefer continuous speech which is fast but it means having a fast machine. This software is more common, with new even-better versions like Voice Express below, appearing almost annually.
After that comes the need for the right gear – and not always the latest gear. Check the advice with the software – the microphone often supplied with it will work, though it may be too flimsy for school use. The small print tells that a ‘USB’ type of microphone gets round a ‘problem sound card’ and that laptop computers often aren’t up to the job. Further advice points to having lots of memory (128Mb say) or the disc will chunter like nuts.
Finally it’s time to enrol by reading, the text of which is a too tricky for some children. You read a very long bedtime story to the machine so it can rough tune itself to the sound of your voice. Over the coming weeks, if you’ve not yet been taken away, you and it will learn to understand each other. You will bond beautifully and get some work done.
FOOTNOTE:
The Government agency Becta together with a variety of schools been looking at this technology and offer user reports and advice through their Inclusion website at