Yes, I know that feeling. Once the concentration starts, it's hard to stop. Perhaps it's a variation on the conservation of momentum: a body in motion stays in motion.
It is pretty amazing sometimes.
I used Dragon NaturallySpeaking 11.5. It's a newer version that has good vocabulary and controls for the things that I do frequently. I think my favorite part of the program is how well it does corrections. There are multiple ways to fix something that isn't right.
My dad actually did research into speech recognition back in the late 80s and early 90s. His Chinese accent was so thick, however, that he broke nearly every system he worked on as well as all his competitors' systems at the annual conventions they had. So I could see why an accent would affect the accuracy of the poor thing.
It was also pretty funny when Dad got into voice identification applications, because he would use my sister and me as test subjects. People often couldn't tell which of us was which by our voices, so it wasn't all that unexpected when the identification systems had difficulty as well.
I'll admit that I also bought the very first DragonDictate and have been following the industry since its inception. At first it wasn't because of my hands so much as being able to use a practical application of pattern recognition research I used to do with artificial neural networks. I remember dad used to use Markov models, but when he saw the results I was getting even on small sets of input, he started to work with some of the same learning mechanisms I had and got really excellent results.
no subject
It is pretty amazing sometimes.
I used Dragon NaturallySpeaking 11.5. It's a newer version that has good vocabulary and controls for the things that I do frequently. I think my favorite part of the program is how well it does corrections. There are multiple ways to fix something that isn't right.
My dad actually did research into speech recognition back in the late 80s and early 90s. His Chinese accent was so thick, however, that he broke nearly every system he worked on as well as all his competitors' systems at the annual conventions they had. So I could see why an accent would affect the accuracy of the poor thing.
It was also pretty funny when Dad got into voice identification applications, because he would use my sister and me as test subjects. People often couldn't tell which of us was which by our voices, so it wasn't all that unexpected when the identification systems had difficulty as well.
I'll admit that I also bought the very first DragonDictate and have been following the industry since its inception. At first it wasn't because of my hands so much as being able to use a practical application of pattern recognition research I used to do with artificial neural networks. I remember dad used to use Markov models, but when he saw the results I was getting even on small sets of input, he started to work with some of the same learning mechanisms I had and got really excellent results.