• rumba@lemmy.zip
    link
    fedilink
    English
    arrow-up
    10
    ·
    2 days ago

    How has speech to text gotten so bad. Dictation and keyboard taps are just going to hell with AI.

    • Duamerthrax@lemmy.world
      link
      fedilink
      English
      arrow-up
      7
      ·
      2 days ago

      The AI bros refuse to even allow proof reading and corrections.

      I’ve never had voice commands work right for my speech. Something I had to demonstrate every time my friends ask why I won’t use the speech commands on their TV.

      Speech interpretation just doesn’t work for everyone and I personally refuse to alter how I speak for a computer.

  • Deflated0ne@lemmy.world
    link
    fedilink
    English
    arrow-up
    96
    ·
    3 days ago

    I believe him, I also believe that Nintendo has secret rooms of deaf kids somewhere. No clue what they’d use them for. But if the news broke that someone found a dungeon owned by Nintendo full of deaf kids i wouldn’t be surprised.

  • lime!@feddit.nu
    link
    fedilink
    English
    arrow-up
    43
    ·
    3 days ago

    it’s crazy to me that for all the ai “advances” in the past few years nobody has thought to improve subtitling.

    • ChaoticNeutralCzech@feddit.org
      link
      fedilink
      English
      arrow-up
      16
      ·
      3 days ago

      Poor deaf kids. Not because they’re being held captive but because they’re relying on shitty automatic captions.
      For example, Czech was only added very recently and the captions really suck, they change the meaning of most sentences and even include spelling errors.

      Everyone making scripted videos should at least:

      1. go through their script to convert it into a transcript (match what’s actually been said – looking at you CGP Grey – and remove visual cues)
      2. upload it for YouTube’s auto-timing (which is not perfect but we’ll take it)

      Too bad the FCC’s captioning act is toothless, even TV stations (like HBO) uploading their content to YouTube don’t bother importing captions even though they’re legally required to.

    • chiliedogg@lemmy.world
      link
      fedilink
      arrow-up
      5
      ·
      edit-2
      2 days ago

      It’s even worse for captions.

      Captions and subtitles aren’t even the same thing.

      In fact, most DVD players don’t even pass the code captioning through HDMI ports, so old captioned DVDs don’t work anymore.

      • lime!@feddit.nu
        link
        fedilink
        English
        arrow-up
        4
        ·
        edit-2
        2 days ago

        explain. edited with explanation. i’ve seen the technology connections video, thanks.

        my comment is still about the actual post above, and i was specifically thinking about auto-generated subs rather than, say, movies. apparently that’s not obvious.

    • Knock_Knock_Lemmy_In@lemmy.world
      link
      fedilink
      arrow-up
      8
      ·
      3 days ago

      Andrew Ng did a video when he gradually added noise to the training audio to improve the quality.

      But here we are dealing with homophones so it’s not just turning speech to text, it also needs to be context aware.

      Possible but too expensive to implement automatically.

        • Knock_Knock_Lemmy_In@lemmy.world
          link
          fedilink
          arrow-up
          2
          ·
          3 days ago

          I’m highlighting that speech to text and context awareness are different skills.

          YouTube is unlikely to waste loads of compute power on subtitles that don’t need it just to capture the occasional edge case.

          • lime!@feddit.nu
            link
            fedilink
            English
            arrow-up
            2
            ·
            3 days ago

            i mean, it’s a one-time-per-video thing. they already do tons of processing on every upload.

              • lime!@feddit.nu
                link
                fedilink
                English
                arrow-up
                2
                ·
                3 days ago

                right now they’re dynamically generating subtitles every time. that’s way more compute.

                • aow@sh.itjust.works
                  link
                  fedilink
                  arrow-up
                  1
                  ·
                  2 days ago

                  For real? That’s incredibly dumb/expensive compared to one subtitle roll. Can you share where you saw that?

    • Ansis100@lemmy.world
      link
      fedilink
      arrow-up
      8
      ·
      3 days ago

      My student friend tells me that the auto-generated captions for non-English MS Teams lecture recordings recently have improved significantly and have even become usable.

    • Vinstaal0@feddit.nl
      link
      fedilink
      arrow-up
      1
      ·
      3 days ago

      Probably because a lot of countries either dub the content or it is already in their native laguage. You generally see a lot of subtitles on OpenSubtitles of countries like The Netherlands where that doesn’t happen

        • vxx@lemmy.world
          link
          fedilink
          arrow-up
          4
          ·
          3 days ago

          Auto generated subtitles don’t sell ads and don’t aquire personal data.

        • Vinstaal0@feddit.nl
          link
          fedilink
          arrow-up
          1
          ·
          3 days ago

          No, but on the fields where there is money to be made for subtitles, like movies and tv shows.

            • Vinstaal0@feddit.nl
              link
              fedilink
              arrow-up
              1
              ·
              3 days ago

              Where do you think money for development of new shit comes from? From places where money is made, like TV, movies etc. YouTube doesn’t really make more money because the ads are there.

              And Twitter hahaha I can only laugh at that

                • Vinstaal0@feddit.nl
                  link
                  fedilink
                  arrow-up
                  1
                  ·
                  2 days ago

                  it’s crazy to me that for all the ai “advances” in the past few years nobody has thought to improve subtitling.

                  That’s what OP say, I just responded with the reason why that wasn’t being done. Aka money