On Fri, May 29, 2009 at 1:41 PM, Quentin Hartman <[email protected]> wrote: > On Fri, May 29, 2009 at 9:42 AM, Keith Lofstrom <[email protected]> wrote: >> But sadly, the only usable speech recognition app >> for her needs, Dragon Naturally Speaking 10 Medical, runs only on >> Windoze, with all the enormous freight of problems that implies. > Agreed. I always think of voice speech recognition as being a very niche > need, but I know a lot of people who would use it if they could afford to / > get it on their platform of choice. There was some OSS speech recognition > stuff that was active in the 90's, but I haven't heard or seen anything > about it in years. As I recall it was based on some code with ancestral ties > to Dragon.
There were a bunch of folks out at OGI (remember OGI?) who were doing "cutting edge" speech recognition work using neural nets back in the early 1990s. It was all using neural nets, and it was mostly funded by one of the phone companies to help automate directory search, etc. I suspect there are oodles of patents to work around for someone who wants to build open source speech recognition software. Keith, how's the performance of it? Is there a genuine opportunity to build something "better" in open source, knowing that one would essentially most likely have to start from ground zero? -- M. Edward (Ed) Borasky http://www.linkedin.com/in/edborasky I've never met a happy clam. In fact, most of them were pretty steamed. _______________________________________________ PLUG mailing list [email protected] http://lists.pdxlinux.org/mailman/listinfo/plug
