For my degree course, I had to transcribe a large number of interviews.
I didn't fancy all of that typing. So I decided to explore speech-to-text technology.
I had tried it before when I had a tablet PC with Windows XP. But the capability was quite limited.
I did some testing with Windows Speech Recognition (on Windows 7) but again the capability was quite poor.
So I bought "Dragon Naturally Speaking" and an inexpensive headset.
I bought the "Premium" edition. Here's a tip: if you're thinking of buying the product then engage in an online chat with a sales person. They may be able to give you a substantial discount off the RRP.
Thus far I have only used it for transcriptions but I find it to be quite effective. Here's how I use it:
- My headset has two jacks (I purposely did not buy the USB kind. I plug the microphone into my PC and the earphones into my Android phone, where I have the recordings.
- I listen to a sentence and then pause the playback.
- I repeat the sentence and check that the program has recognised the words correctly.
The speech recognition works quite well: I can speak at a fairly natural rate, as long as I pause at the end of each sentence.
I did have to "train" the software of course, and it does learn new words so in theory the more I use it the better it will get.
I don't know if this method was faster than typing (although I suspect that it was) but it was certainly more interesting/less laborious.
So far I haven't had occasion to use it for anything else, but I must try to do so.
Some day we will all talk to our PCs/SmartPhones (until such time as we get neural interfaces!)
Here's a video advert for the new version of the product: