Undetectable Humanizer: Lifetime Subscription
Transform AI-Generated Text into Human-Like, High-Ranking Content & Bypass Even the Most Sophisticated AI Detectors
Get 95% Deal

The future of Mozilla's Speech to Text project DeepSpeech is uncertain

Martin Brinkmann
Aug 24, 2020
Firefox
|
32

Following the layoffs at Mozilla and rumor that the organization would be able to extend its search deal with Google for another three years, it became clear quickly that things needed time to settle down.

Employees who were fired would provide insight on how the letting go of employees would affect certain projects at Mozilla. A week later, Reuben Morais published an article on the official Discourse site of Mozilla about the future of DeepSpeech.

future deep speech

DeepSpeech, or Mozilla Voice STT (Speech To Text) is an open source Speech-to-Text engine that is trained using machine learning techniques. It is designed as a counterweight to closed source services operated by Google, Amazon and other companies.

DeepSpeech uses machine learning techniques that are based on Baidu's Deep Speech research paper and Google TensorFlow for its implementation. The service can be run on a wide range of devices in real time including Raspberry Pi 4, devices that run Windows, OS X or Linux, Android, and iOS.

Morais admits in the first paragraph of the post that he does not know how the layoffs and the restructuring of the entire company affects the DeepSpeech project.

Unfortunately, as of this moment we don’t have concrete answers to give. We’re working to find out if the project will have a new home in the restructured Mozilla, and what changes would be necessary for a successful transition.

Morais notes that the team will release DeepSpeech 1.0 soon as "most of the technical changes" landed already and that the team sees "no reason not to ship it". The team will continue its work on DeepSpeech until an official decision has been made and the team is informed about it.

Closing Words

Not all is lost if Mozilla would decide to end support for DeepSpeech. The project is open source and it is possible that another organization or individual developers will fork it to continue development. DeepSpeech is not the only VTT project that is open source either, e.g. Kaldi, Espresso, or Nvidia OpenSeq2Seq, are also open source.

Common Voice, designed to "help teach machines how real people speak", has also been moved into maintenance mode as the future of the project is unclear.

Now You: What would you do, as Mozilla?

Summary
The future of Mozilla's Speech to Text project DeepSpeech is uncertain
Article Name
The future of Mozilla's Speech to Text project DeepSpeech is uncertain
Description
The future of DeepSpeech, a Mozilla Speech to Text project that is open source, is uncertain after Mozilla's layoffs announcement.
Author
Publisher
Ghacks Technology News
Logo
Advertisement

Tutorials & Tips


Previous Post: «
Next Post: «

Comments

  1. NCSA Mosaic said on August 30, 2020 at 4:13 am
    Reply

    What would you do, as Mozilla?

    Sell and/or give away all superfluous asserts, and just work on Firefox and base it on Chromium.

    Some may not like that, but it is likely the most survivable path for Mozilla.

  2. jaap said on August 25, 2020 at 11:02 am
    Reply

    If they advertised common-voice (that collects voice samples in 50 langues) a bit more, provided sufficient hardware for running the common-voice website, they would have enough samples for many more languages by now, which would improve the deepspeech STT enough to build revenue models around. How many browsers have speech-based browsing build-in for many languages, without sending recorded audio to some external server?

    Instead they focus on products that are no different from closed-source providers, and time and time again discover that they are not in a position to compete.

    I don’t really care about deepspeech, it is likely to be replaced by even better systems eventually. But I do hope Mozilla can keep common-voice running, so there will be a fully open (CC0) speech dataset for anyone wanting to implement new or improved speech recognition software.

  3. anonymous said on August 25, 2020 at 4:09 am
    Reply

    Bring back Brendan Eich.

    1. Martin Brinkmann said on August 26, 2020 at 6:59 am
      Reply

      Mozilla would never do that, and Eich would not go back considering that Brave is doing quite well.

    2. Stan said on August 25, 2020 at 4:37 pm
      Reply

      Make it so, worked for Apple, nudge nudge ;) ;)

  4. nealis said on August 25, 2020 at 1:27 am
    Reply

    Wish they prioritized the fundamentals, which is make a competitive browser on desktop and mobile. Even the new Firefox on android is still slower and janky compared to Chrome on android. Everything else like deepspeech is just extra.

    Not that I will personally be around for it b/c I finally uninstalled Firefox from my pc, first time since I can remember I didn’t have Firefox on my personal computers.

  5. Max said on August 24, 2020 at 11:39 pm
    Reply

    What would you do, as Mozilla?

    Fork Pale Moon.

    1. m3city said on August 25, 2020 at 9:13 am
      Reply

      Palemoon is a mess and a relic. The longer it exists, the longer it clinges to obsolete extensions and approach to www. Funny they have continuously readjust it, mimic firefox just to be able to show www the proper way. Forking and forking extensions that are no longer maintained in xul – and when developers cannot be forked then it means they dilute manpower to more and more side projects.

      Used it for years. Checked it just last week. Checked pages that I open daily, and it opened it visually slower, scrolled slower than firefox. But it does work faster on really old, single/dual core laptops and low RAM.

      1. Iron Heart said on August 25, 2020 at 9:33 am
        Reply

        @m3city

        Pale Moon might be a relic, but at least it isn’t spyware that can run arbitrary code behind the user’s back.

        Pale Moon 1 – 0 Firefox

        Not a Pale Moon user btw.

      2. Martin Brinkmann said on August 26, 2020 at 6:59 am
        Reply

        That’s like saying Lynx browser is a relict but at least it is even less spyware. Pale Moon has its fan base, but the browser will have a hard time coping with changes in technology on the Web in the coming years.

      3. Iron Heart said on August 27, 2020 at 9:31 pm
        Reply

        @Martin Brinkmann

        Lynx is a bit of a gross example here, since this browser doesn‘t care for modern web technologies at all, it can‘t be compared to (the maintenance effort required for) Pale Moon. Pale Moon has its weaknesses which I do not deny, but that they do not spy on their users deserves respect in its own right these days.

        Mozilla could totally release Firefox with a similar privacy level to that of Pale Moon, they just refuse to do so.

      4. Iron Heart said on August 26, 2020 at 12:58 am
        Reply

        @Ray

        Firefox has a more modern technical basis than Pale Moon, but I hope you are not really arguing that Firefox is more privacy-respecting, because it isn’t. Most definitely isn’t.

        @m3city

        Firefox is not far away from Chrome by default, and the defaults are the state in which most people use their browser.

        Brave is the most privacy-respecting browser ootb, like it or not:

        https://www.ghacks.net/2020/02/25/study-finds-brave-to-be-the-most-private-browser/

        The “referral link” issue was blown way out of proportion (mostly by Firefox fans, quelle surprise), it was not even a privacy issue in the first place. The referral existed only to let Brave partner websites count the cumulative number of Brave users visiting them (which is why the referral was static instead of randomly generated, if they wanted to identify specific users they would have picked the latter). I have debunked this trumped up allegation here in greater detail:

        https://www.ghacks.net/2020/08/13/chrome-may-soon-only-show-the-root-domain-name-by-default/#comment-4470718

        Brave doesn’t even use referrals anymore upon community request, despite them never having been a privacy issue. But hey, I guess FF fans will be trying to milk this story as long as they possibly can, using misleading information in the process. If you insist on digging out the bad, the worse and the ugly, then here is some neat info for you:

        – Mozilla force-installed an extension that changed the HTTP header without user consent (Mr. Robot incident).
        – They hijacked Firefox downloads with actual spyware without informing the user of said spyware, it did send the browsing history and website interaction data to a third party (Cliqz incident).
        – Firefox Android comes with three built-in trackers (Leanplum, Adjust, Google Analytics) which are non-removable. The user is never informed of this.
        – Firefox’s internal tracking protection doesn’t want to hurt Google and thus whitelists several Google-related trackers.
        – Mozilla can still remotely install FF Experiments as they see fit, they are able to run arbitrary code behind your back.
        – Firefox installs scheduled telemetry tasks that run outside of the browser itself, it collects telemetry about those who have deactivated telemetry and hides away the related about:config entry by default.
        – Mozilla misuses Firefox’s notification system to push political propaganda on users.

        There is more, but I’ll stop there. And you are up in arms about Brave letting a very limited number of known partner websites counting the Brave users who visit them, without even infringing user privacy in the process? Seriously? There are much bigger problems at your own doorstep.

        If those facts I just mentioned are a “crusade” to you, instead of you just accepting them as provable facts as you should, then this says it all.

      5. Martin Brinkmann said on August 26, 2020 at 6:57 am
        Reply

        To be fair, some of the “incidents” that you mention affected an equally small number of users, e.g. Mr. Robots.

        I do agree that browsers need an opt-in for experiments of any kind, regardless of whether they are installed over the Internet or through browser updates. The same is true for Telemetry.

      6. m3city said on August 25, 2020 at 11:19 pm
        Reply

        And FF does? How about chrome, or brave that tinkered with urls, and when caught “sorry, will fix that”. And if PM is relic, then it’s a closed road for continously evolving web, hardware, software.

      7. Ray said on August 25, 2020 at 11:07 pm
        Reply

        > Pale Moon 1 – 0 Firefox

        Hell no.

  6. VioletMoon said on August 24, 2020 at 6:51 pm
    Reply

    It is odd that Firefox used to be the only browser I would use; over that last month, however, I’ve been using Chrome more and more and more. No explanation.

    However, a company that can really make Deep Speech type programs work 99% true to voice, will make billions. Google okay; MS Dictate really shines.

    Maybe its my microphone!

    Even Dragonfly has limitations.

    My father was a stenographer, and to watch him come home every night and read stenography notes and type out a transcript . . . the need for VTT is totally needed!

    I also found that working with students, some students get that “blank page” fright that continues forever. But when I started using Google or Dragonfly with them, they could read like prose [pun intended–not that funny Violet]. It’s because they hated keyboards, or simply couldn’t type; and all that spelling and grammar stuff really messed with their heads. But they had stories to tell . . . .

  7. Peter said on August 24, 2020 at 6:23 pm
    Reply

    This is why Mozilla is loosing; they’re trying to imitate Google by having all this nonsensical projects no one cares about, in a desperate and sad attempt of being the ones pushing the envelope on web technology. You are not going to replace Google and become the foss saviors of the web, Mozilla. Get back to being a browser and stop trying to be Google or deprecate like Internet Explorer.

    1. m3city said on August 25, 2020 at 9:07 am
      Reply

      I disagree on few frounds here. Mozilla was/is the only one that had/has capacity, skill and manpower money to carry on such projects. Monopoly are no good for consumers and freedom of www, and if Mozilla did not try to do it – no one would do it.

      There is so much whining about Firefox, how Libreoffice is crap, how Linux is bad… but all those inspired people to create and share, participate for something greater than money and capitalization.

      And if these projects survive and beceme utilized in usefull applications then it would benefit all of us. I suspect moaners will start bashing such applications then.

      1. Iron Heart said on August 25, 2020 at 9:43 am
        Reply

        @m3city

        > Monopoly are no good for consumers and freedom of www, and if Mozilla did not try to do it – no one would do it.

        OK, so how exactly would they do that? Google owns several exceedingly popular services of the web (Search, YouTube, Maps, GMail, Docs – just to name a few), if Google were to push some kind of web technology which Mozilla doesn’t like, without which those services would break / refuse to work, the chance of Mozilla not adopting that web technology is about 0%. It doesn’t matter if they are using a different engine, Google will get its way, by brute force if necessary. Mozilla can’t and won’t oppose anything.

        By the way, Google does not even have to apply brute force – Google and Mozilla are in a harmonious business relationship, and whenever Google proposes a new web standard at the W3C (a new web standard requires two votes to become official), they can count on Mozilla applauding and voting for their proposal. There is also Apple, but Apple has voted against several of Google’s proposals so far – contrary to Mozilla.

        Firefox is a weak token opposition that is required to keep the de facto Google monopoly going, and to avoid antitrust scrutiny. Google has no desire to be broken up, you see.

        That you think Mozilla is “opposing” anything (which they can’t, even if they hypothetically wanted to) is deeply illogical, and goes to show that Mozilla’s marketing, misaligned with their actions as it is, is still extremely effective. Thanks to the existence of Firefox, which conveniently absorbs all the anti-Googlers into the Google camp, real opposition will never be able to rise.

        The browsers Google hates the most: Safari, Brave (since it attacks their business model), several minor ones like Pale Moon or Falkon.

      2. m3city said on August 25, 2020 at 11:17 pm
        Reply

        I have a feeling that you have no idea what a monopoly is, and what the long term consequences of monopolies. I come from a country, where a communist goverment had monopoly to everything, politically, socially, economically -wise. And I saw what it lead to.
        I will just say that my understanding of google monopoly in a capitalism world is very close to one my country experienced through communism, with a more subtle methods of course.

        And Mozilla or any other company does not have to oppose. They may promote, inform, propose new solutions. One may say, Google can do this too. And thats true, but the motives count. You see, the only motivation, moreover – one that will nullify any other when this motive is in danger – is income. And I… opose that.

        Hate between google and others? Come on, man. Your crusade-style flame wars here on ghacs are getting funny.

      3. Iron Heart said on August 26, 2020 at 12:36 am
        Reply

        @m3city

        Your response was totally nilly willy, as expected. The fact is: Mozilla can’t do anything against Google, even if they wanted to, because Google also owns several popular web services and can push through their ideas using those. You using Firefox doesn’t achieve a thing. You brought up communist dictatorships; well, in such a scheme Mozilla would be akin to a fake opposition collecting real dissidents and later turning them in. No real opposition to be found there, until something else arrives.

        > They may promote, inform, propose new solutions.

        Ineffective. Google would have to be broken up for real change to take place, most likely. But them being broken up would require a successful antitrust case, the existence of Firefox prevents exactly that from happening.

        > Hate between google and others? Come on, man.

        There are definitely browsers Google dislikes, Firefox is not one of them.

        > Your crusade-style flame wars here on ghacs are getting funny.

        Maybe, but as long as you fail to raise a valid counterargument I guess my drivel is not all that pointless, after all. :D

      4. Martin Brinkmann said on August 26, 2020 at 7:04 am
        Reply

        Mozilla sits in a precarious position. It relies on Google money for the most part, and that money is paid because Google can display ads when Firefox users search using the search engine.

        I would bet that the contract between the two companies states that Mozilla cannot implement technology into Firefox that would prevent Google ads from showing. Or, Mozilla is paid based on search performance and not a lump sum.

        Mozilla executives decided that the best way to go forward is to establish other revenue sources to decrease the reliance on the search deal. That’s definitely an option, but it is not clear how successful it will be.

        A more risky approach would have been the following one: agree to a partnership with a privacy-first search engine. Yes, these cannot pay Mozilla as much as Google, probably, and they don’t have the cash to pay in advance, but both parties could reach a revenue share agreement or something. Doing so would improve Firefox’s standing in the privacy community and probably also for other users, and it would remove the dependency on Google when it comes to financials.

      5. Iron Heart said on August 27, 2020 at 9:28 pm
        Reply

        @Martin Brinkmann

        Your assumption is most likely correct. One has to wonder why Mozilla fails to implement and / or enable some privacy improvements, despite them being non-breaking in nature. Examples? Why doesn‘t Firefox com with an ad- and tracking-blocker that is not a joke, why allow prefetching, why enable some fingerprinting friendly APIs that have little to no use case in the wild, who needs resource timing stuff etc. etc. Contrary to what some people believe about me here, I am NOT blaming Mozilla for having to implement some stuff without which Google web services (and others using those technologies) would break; they have no choice in cases like this. However, I have a problem with them…

        – not refusing to implement or default-disable some stuff that we could easily do without and
        – them claiming that they are oh so independent when reality paints a much different picture, this is just misleading.

        I guess Mozilla teaming up with someone like DuckDuckGo would mean a healthier influence than Google, that‘s for sure. Not sure though whether or not DuckDuckGo has the financial capacity to fund Mozilla, even if Mozilla were to reduce themselves to development of the core product (Firefox). DuckDuckGo is just too small still, IMHO. It would be possible for DuckDuckGo to team up with Mozilla on a smaller scale though, perhaps in select regions.

        As I hinted in my comment above, real change would require Chromium development to be wrestled away from Google, perhaps by Chromium being transferred to an independent foundation governed by multiple companies. The combination of having the dominant browser and popular web services ain‘t healthy.

        But then again, we are still better off than we were in the Internet Explorer days – one can take the Chromium code and modify it to suit specific needs. So far, Brave seems to be able to design a privacy-first browser successfully, Vivaldi seems to be able to design a customizable browser successfully etc. What could be problematic would be Google introducing big enough changes that those become impossible. Compared to Google, Brave Software and Vivaldi Technologies are minor companies and minor contributors to the Chromium codebase. The reversal of potentially damaging changes they can pull off have their limits… So in summary, I think transferring Chromium to a neutral foundation would solve a lot of issues, but considering that the minor Chromium forks can achieve their goals just fine so far, I think the current situation is not really concerning.

        What people here refuse to get about me here, apparently – I have a realistic outlook on Mozilla, the implications of the Google search deal, and their de facto capability to oppose Google which isn‘t noteworthy. If they were at least trying to improve user privacy in areas where it‘s doable, that would be a saving grace, but since they refuse to do that I have no sympathy left for them. I take a far less problematic Chromium fork over the duplicitous token opposition, thank you very much. I realize that this is highly controversial here and also a minority opinion, though. If you ask me, many people here have extremely unrealistic expectations towards Mozilla as far as their „anti-Google potential“ (if you will) goes, yet paradoxically cut them a ton of slack in areas where they could probably do something to improve user privacy, but don‘t… I don‘t get that at all, but anyway, not my cup of tea.

  8. Anonymous said on August 24, 2020 at 5:15 pm
    Reply

    What would you do, as Mozilla?

    1. Appoint an adequate CEO.
    2. Do everything to improve Firefox.

  9. Stan said on August 24, 2020 at 5:04 pm
    Reply

    92 out of 100 users use browsers other than Firefox, why is it necessary to have a ‘news bulletin’ every time someone at MozCo breaks wind?
    Just asking…..

    1. anona said on August 25, 2020 at 10:41 am
      Reply

      There’s only so much news you can bring about Chrome, Edge and Opera.

    2. Iron Heart said on August 25, 2020 at 12:03 am
      Reply

      @Stan

      IMHO, @Martin Brinkmann is in a pretty tough spot. The last decade of Firefox’s history can be summarized as a colossal failure, its market share sharply declined from about 30% (2010) to about 5% overall (2020). Since this trend is continuing with no sign of reversal, it doesn’t exactly take Albert Einstein to realize that Firefox – as a topic – is slowly fading away. However, this blog also has a substantial Firefox audience. So, on the one hand the software is slowly declining, on the other hand @Martin Brinkmann has to cater to his readers… As I said, tough spot to be in, but I wouldn’t be surprised if Martin already had some kind of backup plan. Perhaps this blog will become more operating system-focused, or perhaps attention will shift to various browsers that are currently underreported. I could also see extension-related topics dominating. Anyway, Firefox is not what you would call a long-term plan anymore, that’s for sure.

  10. Niko said on August 24, 2020 at 3:54 pm
    Reply

    750+ employees and can’t even maintain a web browser, not to mention a few extra projects on the side! Sad what mozilla has become …

  11. ULBoom said on August 24, 2020 at 3:45 pm
    Reply

    TL;DR
    DeepSpeech: TBD.

    Wow, someone in a Big (sort of) Tech company who can write coherently; no use or abuse of “experience” anywhere.

    We’re cookin’ now!

    Sad Google Serfdom’s products matter so little.

  12. ShintoPlasm said on August 24, 2020 at 2:37 pm
    Reply

    Another waste of time that should never have been started.

  13. yo said on August 24, 2020 at 2:13 pm
    Reply

    No shit! they used all their money on diversity and fired actual programmers, even i, a middle eastern disgusted by this.

  14. Iron Heart said on August 24, 2020 at 1:03 pm
    Reply

    > What would you do, as Mozilla?

    Refocus on, you know… Firefox, maybe? I’d mainly focus on dropping all nonsense projects (like the one described in the article) and restoring sanity to Firefox first. This is how I would fix Mozilla (extended):

    https://www.ghacks.net/2020/08/14/google-and-mozilla-to-extend-search-deal-according-to-reports/#comment-4470774

    The realistic outcome though, would be them switching to a Chromium base in the near future, proceeding to use nuFirefox as a mere vessel for paid services. While this won’t stop the market share bleed by any means, it allows the Mozilla management to enjoy their Google-sponsored salaries even longer, with only a minimal developer crew required to keep this Chromium fork going. You don’t believe this? Well, ask the former Servo team…

Leave a Reply

Check the box to consent to your data being stored in line with the guidelines set out in our privacy policy

We love comments and welcome thoughtful and civilized discussion. Rudeness and personal attacks will not be tolerated. Please stay on-topic.
Please note that your comment may not appear immediately after you post it.