First Foray Into Speech Recognition Software
I’ve been using the Motion Computing F5 Tablet PC in conjunction with Dragon NaturallySpeaking software over the last few days as the novice-beginner user in John Gannon’s Speech Series. The training program was very easy and allowed me to get used to the program as well as train myself in how to speak to a computer, which surprisingly is very different than speaking to people. I have actually found that using the speech recognition software has both benefits and drawbacks. Since the motion tablet PC does not have a keyboard I am actually using the Dragon NaturallySpeaking software to write this blog by dictating my words instead of typing them or writing them. Unfortunately I am bound by the motions built-in microphone arrays since there are no ports on the tablet for peripheral microphones. The errors that may occur during my dictation is a result of using the built-in microphones on the motion tablet rather than in the major issue with Dragon software. I will leave the errors included in this blog post Sluggy can see how well the Dragon NaturallySpeaking software actually does on non-ideal equipment.
John Gannon has already spelled out the minimum requirements for Dragon to properly work on a computer. The F5 I am using only has a 1.2 GHz single core processor and 2 GB of RAM which is well short of what is optimum for Dragon to be fully efficient. When Dragon NaturallySpeaking is running there are definite slowdowns in computer functionality. Sometimes after speaking longer sentences have to sit there and wait for Dragon NaturallySpeaking to actually catch up to what I am saying. Another thing I’m trying to get used to is actually saying aloud the punctuation that I am looking for. Otherwise using Dragon NaturallySpeaking software has actually improved the functionality and usability of the motion tablet without connecting to an external keyboard. So long as I use the software in a quiet environment with no extraneous background noise, it seems to do very well and is very accurate.
However if I am near a computer that has a dish and horrifying China do some blogging while watching TV the volume with either they or the errors increased dramatically. This last sentence was actually said in front of the TV on lower volume that I can hardly hear but apparently the microphone can pick up the background noise which throws off the speech recognition software. Having even slight background noise decreases the accuracy of the speech recognition software greatly. I can see how using a dedicated microphone that has decent background noise reduction can really make the Dragon software exceptional. It has taken me 10 minutes to write this blog by dictating it. If I were to have tried to write this blog by using the tablet ink function it would’ve taken me at least 10 to 15 minutes. I think if I were to use this tablet PC more often and get used to the speech recognition software as well as the speech recognition workflow, using speech technology can speed things up quite a bit. Another thing I’ve noticed about the Dragon NaturallySpeaking software is that the more I use it the more accurate it gets.
Over the next few days they’ll be trying to use the Dragon NaturallySpeaking software to navigate the Internet through both Internet Explorer and Firefox. I’ve given it a try recently and although things do work still feels a little cumbersome because I’m not used to the specific software commands as well as the navigational voice commands to make the browsing experience actually enjoyable. As of right now using Dragon NaturallySpeaking to navigate through Internet Explorer or Firefox seems to be a little cumbersome and counterintuitive. Stay tuned as I continue to use this awesome software and hopefully I’ll be able to fully utilize all of its features and benefits.