Hi Guys,Now that the 508 refresh and NFB convention are behind us (look for a gonzo report on the convention is a week or so on BlindConfidential) and we now have an interesting band of volunteers from around the world, we (FSF/GNU) will be kicking off some of the initiatives we' have been talking about since I joined up.
The first major task regarding dictation software and adapting for people with a wide variety of use cases is to come up with two major tables: first, what is needed by the major serpopulations (this should range from medical doctors during surgery to quads who can talk but not move or any thing in between). Second, we need to assemble a team to start looking at the five or six floss speech recognition programs and correlate what's been built with what we need and assess quality to ensure that our initial choice is scalable enough that we can add features and fix bugs to build a valuable tool for our user community.
This research will require some "heavy lifting" and will be pretty boring at times but we need to remind ourselves that the community of free software hackers is small and we need to optimize the effort. Working on multiple speech recognition programs is nothing more than counterproductive sectarianism. We can take the best ideas from each, and in some cases, code too to work on this effort.
We have set up a mailing list, speech-reco@xxxxxxx (join by sending a blank message to speech-reco-request@xxxxxxx with subscribe in the subject line) this list intends to bring the ideas together and makes it possible for people to discuss the various merits of the software we are evaluating and, at the same time, talk about which of the existing programs address which use cases from our other table).
FSF/GNU has no horse in this race, some of us may have opinions but we need to work together to make a solution about which we can be very proud.
Also, if you haven't yet, send an email to accessibility-request@xxxxxxx to subscribe there and you will be able to keep up with other free software accessibility initiatives. Project GNU want to do our best to facilitate each step of the process but the population on the speech_reco are going to be calling the shots. I am a very good resource for management tasks but, other than having used dictation software in the past, I can add little value to the project details beyond keeping track of some important items and working with the media to make sure the world hears of our effort and that volunteers know how to find us.
HH, cdh 07/2010 03:04 PM, Bill Cox wrote:
Hi, Stefik. On Wed, Jul 7, 2010 at 11:52 AM, Andreas Stefik<stefika@xxxxxxxxx> wrote:I'm not an expert on voice input for programming, but I know the researcher Andrew Begel did some work on this a few years back in Java. Any connection here in any of these tools?He worked on some cool stuff while at Berkeley, but like many university projects, his Spoken Java does not seem to have gained traction. Andrew now appears to work for Microsoft, and hopefully he can contribute to accessibility there.We've also considered adding voice support into Sodbeans, our FOSS tool for blind users. If we wanted to hook in to experiment with, how difficult is it to incorporate some of these existing tools?The current tools, VEDICS and Simon, are at a state that isn't really quite usable, but which show clearly how close they are. I would guess that it's a good time to get programmers with typing impairments involved, so we can guide these projects in a good direction. I would wait until there is a bit more infrastructure in place before hooking into it. Bill __________ View the list's information and change your settings at //www.freelists.org/list/programmingblind
__________View the list's information and change your settings at //www.freelists.org/list/programmingblind
