![]() |
KnowBrainer Speech Recognition | ![]() |
Topic Title: Microsoft debuts "Voice Access" feature in Windows 11 Topic Summary: Created On: 12/10/2021 11:42 AM Status: Post and Reply |
|
|
![]() |
|
FYI, Microsoft is introducing a new speech recognition interface right alongside the current Windows Speech Recognition implementation. It's called "Voice Access," and it is currently available in a Windows 11 development build. Here is a blog post about it. (Scroll down to the "Introducing Voice Access" section.) Announcing Windows 11 Insider Preview Build 22518 | Windows Insider Blog
|
|
|
|
![]() |
|
Thank you - this is really interesting. They have provided interesting features and quite a long list of commands.
------------------------- Writing and editing (my main website): Welcome - Words for Sale The woman who dueled with Aaron Burr and won: www.MmeJumel.com Crohn's News Blog: www.crohns-news.net |
|
|
|
![]() |
|
"Voice access supports English-U.S. language only". Presumably not British English.
|
|
|
|
![]() |
|
Our understanding is Voice Access currently only supports US English but this will probably change when it is officially released. Only the Dev channel (latest features but not as stable) can currently test Voice Access. Beta testers (stable but not finished) will get it later and then the public. Keep in mind that Microsoft probably doesn't want Voice Access to be much better than the Dragon Home Edition, which it already looks dangerously close to. When Voice Access is released, Nuance really should drop the Home Edition because it will potentially be an embarrassment at that point. ------------------------- Change "No" to "Know" w/KnowBrainer 2022 |
|
|
|
![]() |
|
Many have written off Mouse Grid. But here it is presented to a new generation of voice people.
|
|
|
|
![]() |
|
Microsoft is also including a numbering scheme which similar to Click by Voice but works everywhere. We can't think of a way that the Microsoft Mouse Grid wouldn't be significantly slower than using the new numbers.
------------------------- Change "No" to "Know" w/KnowBrainer 2022 |
|
|
|
![]() |
|
It seems like there are now 3 different versions of dictation built into Windows: Windows Speech Recognition, this new Voice Access, and whatever you call it when you press Windows + H on your computer.
------------------------- Dragon Professional Individual v15.6. Windows 10. Knowbrainer 2017. |
|
|
|
![]() |
|
When Microsoft introduced the Edge browser, they kept Internet Explorer a long time. We would like to see Microsoft remove WSR to eliminate possible confusion but our best guess is that they will keep it around like they did Internet Explorer. We further suspect that when you try to open WSR, you will be prompted to use Voice Access instead.
Windows 11 already includes cloud dictation which is equivalent to Voice Access without command capabilities (same speech engine). Bottom line: You are correct. Keep in mind that when you press {Win+H} you are only using the cloud version. When Nuance releases Voice Access, you will have the whole enchilada. We haven't performed any serious testing but we found Voice Access accuracy to be on par with Dragon. However, from a functional point of view, comparing Voice Access to Dragon would be equivalent to comparing WordPad to Microsoft Word. The only Voice Access advantage would be the new Numbers Mode which we expect to appear in Dragon 16 ------------------------- Change "No" to "Know" w/KnowBrainer 2022 |
|
|
|
![]() |
|
I tested Voice Access in one of the Windows 11 beta versions. It's quite accurate but text deployment is a lot slower than Dragon®.
I was pleasantly surprised by the good-looking mousegrid and the fast responding Show Numbers feature (which now are flags).
I was interested if Voice Access dictation would be context aware in some way. That appears to be so, partially but only in Microsoft Edge. You can select a word by voice and then dictate something else. However what you then dictate always appears capitalized.
Nonetheless Voice Access can be a very interesting feature to use alongside with Dragon®. ------------------------- Turbocharge your Dragon® productivity with 40 Power Addons |
|
|
|
![]() |
|
"equivalent to Voice Access without command capabilities (same speech engine)."
What speech engine would this be? I assume it's not the Dragon speech engine they just acquired? what are the differences between the WSR numbers mode and the voice access one? |
|
|
|
![]() |
|
WSR numbering covers the entire control, and sort of blinks off and on to allow you to see the control again. Voice Access emulates the way VoiceComputer has been numbering controls for years. So, I view Voice Access as a scaled down combination of Dragon and VoiceComputer. Given that Microsoft is vastly superior in resources than either of these companies, I would hope they will continue to develop a much more robust and useful product in the long term than what we have available today. But I suspect it will take a long time before that happens. ------------------------- Tom
Programmer Of SP 7 PRO (speechproductivity.eu) |
|
|
|
![]() |
|
Hey Tom, would you mind telling me a bit more about your perspective that voice access is a "scaled down combination"? Are you specifically referring to the number overlays with that sentiment, or speaking about the experience more generally? When you say "much more robust and useful," what are some examples of things you think could/should be improved?
I actually lead the design for voice access, so I'm super curious to hear more. We are always looking to understand what our users like and dislike (or want to see) about the experience, to help us better meet user needs! |
|
|
|
![]() |
|
We would like to jump in if you don't mind… This following is only our personal ($0.02) opinion:
We believe a full function speech recognition program should include a Vocabulary Editor that end-users can minimally add or delete words and the ability to create rudimentary commands; perhaps step-by-step and boilerplate text. Of course this could compete with Dragon which might be counterproductive. Voice Access accuracy seems to be excellent. Note that we are considering porting KnowBrainer 2023 to Voice Access. And… as you can see, “Voice Access” is already in our vocabulary ------------------------- Change "No" to "Know" w/KnowBrainer 2022 |
|
|
|
![]() |
|
After playing with Windows 11 for 1 1/2 days I've discovered the following in the UK version:
SendDragonKeys does not work in advanced scripting (most of my commands in KnowBrainer so I did not notice this at 1st) Voice typing currently does not accept any dictation commands other than stop dictating (apparently this is USA only at the moment) ------------------------- Thanks Mark
Dragon Professional Advanced Scripting/KnowBrainer Scripts |
|
|
|
![]() |
|
I've just had a Windows 11 update which introduced Notepad with Dark Mode. ------------------------- Thanks Mark
Dragon Professional Advanced Scripting/KnowBrainer Scripts |
|
|
|
![]() |
|
I discovered that my Notepad is not functioning well as well . It behaves exactly as yours In addition : -Some dictation meant to be in a new line goes into the same previous line as it it were a continuation. -New lines are not correctly capitalised and they start a tab or 2 inwards as if they are indented I am dissapointed as I like Notepad a lot for doing my dictations
|
|
|
|
![]() |
|
Same here. This is what I get in the latest Notepad (11.2112.32.0) version:
"New Line" voice command has become unreliable.
Notepad is throwing in unnecessary initial and trailing spaces all the time (sometimes even spontaneously trims the spaces).
It frequently does not capitalize the first word of the sentence.
Dictating "." gives erratic behavior.
It sometimes spontaneously inserts random letters in my dictation.
The system menu is not responding well to maximizing and minimizing and or "click close" command. ------------------------- Turbocharge your Dragon® productivity with 40 Power Addons |
|
|
|
![]() |
|
I have been away too long… ------------------------- Tiger Feet |
|
|
|
![]() |
|
Hi there! In terms of what kinds of things I would love to see Voice Access be able to do, please have a look at this amazing little program: https://www.voicemacro.net/ I own a licence for the latest version of Dragon (which I don't use), and have played around with every dictation solution available on Windows (Dragon, WSR, Kaldi, etc.), as well as every voice command/scripting tool (KnowBrainer, Vocola/Unimacro, Talon, Caster/Dragonfly, AutoHotkey, VoiceMacro, etc.), and currently use VoiceMacro in my daily work as a patent translator. It's surprisingly powerful. I use it mainly for commands (to do stuff in my translation software), but also use it for occasional dictation (which I can also start by voice). However, it uses the old-school WSR engine, so is not great at dictating flowing text. I'm hoping the developer of VoiceMacro switches to the new Windows 11 Voice Access dictation technology (if possible), to make it even better. But in terms of what kind of things I would love to see Voice Access do, VoiceMacro pretty much nailed it. Michael ------------------------- Dragon Professional Individual 15 + Vocola + Speech Productivity
|
|
|
|
![]() |
|
I'm currently testing it again, after joining the Insider Beta channel. Select 'n Say functionality, or whatever it's called in Voice Access works nicely in e.g. Notepad and Word, but not (yet?) in my translation software, such as memoQ and Trados Studio. Does anyone have any idea if they will add more apps to the list with select X/Y/Z functionality?
------------------------- Dragon Professional Individual 15 + Vocola + Speech Productivity
|
|
|
|
FuseTalk Standard Edition v4.0 - © 1999-2023 FuseTalk™ Inc. All rights reserved.