Speech recognition software: "Dragon" it out...

After watching all of the glowing reviews for Dragon on television a couple of years ago, I purchased a copy in an effort to increase my productivity. I spent about 30 minutes reading a Dave Barry essay (the only part that I enjoyed) to accustom the program to my voice per the directions. When I tried to use the application, though, it just spewed out gibberish no matter how distinctly I spoke and I gave up. Has anyone tried anything better?

I tried a speech-recognition program a few years ago with similar results. However, it was more than just the inaccuracy of the software that was an issue for me. Writing and speaking don't necessarily rely on the same neural pathways and I just found it much harder to compose the same types of sentences and paragraphs (and to edit them as I write) that come so easily to me when I'm typing. In my opinion, those things might only be useful for people composing bullet points or very simple thank-you notes and things like that than they are for more complex types of writing. Then again, it might also be particular to the person.

This has been my experience as well and I had planned on using Dragon mostly for quotations from print materials that weren't available in digital form until I became sufficiently expert with the technique to use it for composition (just like in the commercials!). Strangely, my younger daughter was able to get more accurate results from the version that I trained with my voice (still far from perfect though).

While speech recognition software still needs some work, text-to-speech has become much better. For example, the service provided by Text-to-Speech Reader at ttsreader can speak in UK or US English (fairly convincing) as well as French, Spanish, Italian, German, Japanese, Korean (I think) and Mandarin (I think) at different rates. This application can be used to listen to a book while you're driving (in case of an accident, you didn't hear this from me).

It's been a while since anyone asked, so I'd like to know if anyone has recently tried and what are your views now? Have the programs improved? Recommendations? I'm a fast typist but my brain is even faster. ;) Side note: I used to see commercials for Dragon a lot but haven't in quite some time, which tells me its popularity may have waned. ??

I've been using Google Voice for a period of time and I have found that it is easier to train this program to understand one's speech patterns.

It pretty much works well except when you go for really complicated words that, when pronounced, seems to confuse the program. In which case, some retraining for the program is required. Usually by updating the dictionary content.

Overall though, I find it useful only for drafting my papers and other documents. When it comes to the finalization of the presentation, nothing beats manually pounding away on the keyboard. If you are also pressed for time, it is still better to just type out the document than speak.

Somehow, it is easier to hit the delete or backspaces button when compared to all the steps one has to go through in deleting a dictated text.

I found myself still having to manually take over at certain points. At this point, I still believe the speech recognition programs still have some way to go before it can be called a true "speech recognition program".

There are still eccentricities and unique patterns in the way that individuals speak which cannot be understood, adjusted to, or overcome by the programs.

I've been using Live Transcribe lately and it has been a wonderful experience. The program itself takes wonderful dictation when required and accurately takes notes like a secretary in a group meeting setting. It is a necessary weapon in today's work from home arsenal. It can really take a badgering when it comes to group discussions. The program itself is not complex and pretty much works as described, just hit the record button and you are done. These programs are also making it easier for students to take notes in class. Cursive? Print writing? What's that? Writing is now a part of the history of man, only to be seen in museums. Voice dictation is the way of the future.

Speech recognition software has come such a long way already. Siri takes better instructions now. Alexa can almost guess what it is I want it to do even before I complete my sentence, Bixby recognizes voice imprints more accurately as well. I haven't had to use my laser pen on my laptop or mobile phone for about a year already. The addition of AI to the speech recognition software has made it the almost perfect digital assistant. However, it still gets confused by some dictated words and needs help in understanding some localized words. I guess that is the going to be the continuously evolving side of speech recognition and AI programs going forward.