This topic obviously has vast implications for human communications, and I will be writing more and more about it. But for now, I have an immediate need for Speech Recognition Technology (SRT) in the classroom that is specific and modular.

Accurate SRT should be easy today, and as a communications teacher, I have indeed noticed that it has improved remarkably even within the last several years. But as both a communications instructor and a Web content creator, what I want is a WordPress plugin that enables me to use speech recognition globally throughout my site in the same way that I can use images or audio clips now. What I would like the plugin most for is testing: I’d like to give students an environment where they are unable to use any input device besides a microphone. Their voices would then be transcribed into text fields which could be automatically graded for errors. In this vision, students would be allowed to try again as many times as they like until they got the correct answer, or move on if they prefer and accept the error. They would not, however, be allowed to copy-paste. Of course, their nearest neighbor who happens to be good at dictation software could input for them, but I would consider that a valuable experience as well. These points are all moot, however, as no such plugin exists and does not appear to be materializing soon. There is one called ‘Dictation,’ but its use is quite limited. I haven’t tested its accuracy and I have no idea how to design one.

On our phones, both Baidu Translate and Google Translate do fine jobs at SRT transcription, translation, and target language reproduction. I do see that those two particular SRT engines are already being implemented in recent apps so I assume that those SRT and TTS engines are free so that in implementation they are close to Open Source. Looking online, I also see many Open Source engine options (over ten major ones) available for the tech-savvy, but I haven’t tried any of them yet. I have many reservations about becoming tech-dependent on huge corporations like Google and prefer open domain resources as much as possible. Parts of their platforms may be free today, but down the road, Open Source is the only real solution in making knowledge and technology free and available to everyone. As an example, I look with horror at how SEO Yoast is now squatting in my private site, dictating terms and image sizes. Still, if I could get an SRT plugin to WordPress today with a Baidu engine, I’d gladly adopt it!

I have a strong interest in tech development, but I couldn’t code my way out of a PHP stylesheet crisis if my life depended on it. While I’m enamored by the notion that “Code is Poetry,” I’m no poet and I know it. So I won’t be compiling the WordPress Robust SRT plugin anytime soon. I sure hope that someone else does though!

One thought on “Open Source Speech Recognition Technology

Leave a Reply

Your email address will not be published. Required fields are marked *