but why is it even happening??
Hey, Mozilla. Here’s a thought. Is there any way that you could make an offline version of the Firefox Voice and Firefox Voice Fill extensions? In other words, is there any way that you could possibly modify the extensions so that they can be powered by users’ computers instead of the servers? Maybe you could combine them into one extension, and modify it to be able to run locally on a user’s computer? Thanks. God bless. Jesus Christ loves you - He always has and always will, no matter what! You’re priceless and you matter!
Keep being awesome! <3
Unfortunately, this is not feasible quality wise, given that even with popular projects like Common Voice and Deep Speech, we never got enough representation to train large speech-to-text models as good as the online ones we used for Voice Fill and Firefox Voice to cover the whole diversity of our user base.
Thanks for your enthusiasm though and for being a user through all this time.
It’s a real shame that this is being retired but I understand having to reprioritize. I was hoping we could add mozilla deep speech is a backend to the dragonfly project but without community support/Mozilla it seems unlikely.
I work with products such as Caster which builds off of Dragonfly framework. The goal of these frameworks is to allow people to control their compute environment by Voice. This allows people disabilities interact in a way of the computer that they would not otherwise be able to or for developers to augment the workflow. We continue to push for developing open source solutions for voice recognition stack.
Part of an open source solution Kaldi https://github.com/daanzu/kaldi-active-grammar where we are developing models for voice recognition commands and dictation. The pre-trained model general English Kaldi nnet3 chain mode is trained on ~3000 hours of open audio. Perhaps that’ll be helpful those who rely on Firefox voice.
Well i remember trying to train Dragon Dictate many many years ago teaching it to recognize voice patterns etc, it was a nightmare. Well i thought that Mozzilla was in the ballpark with creating a similar open source voice input although limited, voice recognition process that required minimal if no training and having tried Voice Fille and Voice Beta extensions, appeared to have ironed out many of the annoying stinks out without the requirement and needing to sit and read the Art Of War or Shakespeare before the progrann anaged to get the first line errect.
Oh well again the onset or ability to use ones voice too fill out forms / transcribe text and incubate or troll our Hech without need of hours ttwo get it to orgnize a sentence without having to redit and retype the (d)occumentevery second word was either not understood, or the (s)oftware would make up its own mind.
It is funny that although we now carry mobile phones - ring ring “hello” - to the dream was walk to reach others essentially, (b)ending and §yping text messages, like another antiquated technology the pager.
The above mistake were int(er)entional
I hate say i am old enough - just - remember writing a letter and posting it. Then came the mobile brick, but for some reason just like the lost concept of talking to each other - watching people at the same table texting each other and except for yes or No a giggle between sending an sms to the person next to you or across the table has become the norm.
As someone who has difficulty using a keyboard and know people who would find the use of voice commands, would improve many lives i find it strange that having used Dragon Dictate 10 or more years ago, we still sit and plonk and thud a keyboard or phone screen, the ability to speak or hold a conversation has discapeared if born after 2000. However being able to control ones tech with voice commands or dictate a letter is still limited and software etc, is often expensive but when it works can make a long painful job a little less expensive and a deal a lot less annoying.
I must say i will miss Voice fill and Firefox voice beta will be a loss for may of us that find it difficult to use a keyboard etc. But it seems very clear that voice recognition, dictation or simply accessing or commanding our tech via voice is something that just either is yet to catch on or as many have found the software is still expensive and
requirement for training seems to have been stuck in the too hard or not viable . Although i have found Mozilla’s foray into this area of tech basket and buried deep in the internet cemetery. Although not functional full hardware It will be a loss. The glimmer of hope is with Mozilla being open source someone may pick it up in i hope not to distant future and will see it as challenge and pick it up again and use them as a template too develop this software further.
Discussion in Reddit: https://redd.it/l81odx
… no explanation for the decommissioning. I can guess a reason, but I’d prefer the explanation to come from Mozilla.
Primarily economics, I guess. Plus maybe logistics, given COVID-19 restrictions. @caitlin please, can someone at Mozilla confirm?
In any case:
… Mozilla has realigned its priorities, and Firefox Voice will no longer be supported by Mozilla. The Firefox Voice team will be going to new things, no longer with Mozilla. …
Changing World, Changing Mozilla - The Mozilla Blog (2020-08-11)
Test Pilot/Voice Fill category (archived)
@LexiconCode please, can you suggest a forum that’s suitable for discussing a range of possible solutions? Not Firefox-specific.
For what it’s worth, https://www.google.com/search?q=“Speech+recognition”+“open+source”+forum&tbs=li%3A1#unfucked led me to (amongst other things):
– running phpBB. I’d prefer a Discourse forum but I shouldn’t be fussy.
I don’t have any recommendations for a forum at this time. Most of the work I do is coordinated through chat via Gitter/Matrix. However a Forum format would be preferable for such a discussion.
I took up to FF because Chrome is a HUGE Resource Hog on my system…
I mean … it slows to a crawl if I have a number of tabs open… gets unusable.
so I won’t be going back to it any time soon… for sure…
Thanks for replying. I understand. It’s unfortunate that Mozilla can’t continue these services, but I’m proud of Mozilla for being a privacy-conscientious, user-focused company. Thanks for all that you do! May God bless you. Jesus Christ loves you - He always has and always will, no matter what! You’re priceless and you matter!
Keep up the great work!
Kind Regards, SubjectTea9
I have Dragon NaturallySpeaking Pro. I have seen that outperformed by Speechnotes. I don’t fully understand this move and seems to be in the opposite direction from what’s trending. I hope you guys reconsider. Thanks.
I had to reply to you because I couldn’t figure out where to post ;-(
Hi. I just looked into Speechnotes. I see that it requires you to pay through a Google account, is it compatible with Brave? And will it stay compatible with Brave after Google makes those changes to the chromium APIs? There are some use cases where it would be helpful for me, but it’s not worth it if I’m stuck with the bloat and user tracking of vanilla Chrome.
Yes as a long term user of firefox on Mate, I’m sad to see yet more areas of this browser made less, not more, user friendly. Up to now I only had to switch to Chrome for casting.
sad sad day to see removal of voice
Hello, Isn’t the server side open-sourced as well ? that would allow users in nee of this functionnality to run their own stack
Thank you for the information… what does an ordinary non-computer person need to do… can I still voice on zoom and stuff like that… if so I am a ok…
I would also like to be able to do voice searches on Firefox ! Google chrome does this very well already, but I prefer Firefox.
Hi @bj_johnson2003, this should only affect your ability to do voice search on Firefox. You’ll still be able to use voice features for other products like Zoom.
I run my fashion blog so I research new topics each day which makes me tired sometimes. But with Firefox voice recognition I enjoy researching. Why retiring voice fill and firefox voice extension
… we are decommissioning the addons purely because the use of Google’s STT backend won’t scale for us, and we never managed to train production ready models through projects like Deep Speech and Common Voice with the quality required for VF and FxVoice.