• thirteene@lemmy.world
    link
    fedilink
    English
    arrow-up
    25
    arrow-down
    2
    ·
    10 months ago

    I’ve been watching a few projects that are attempting to live translate videos. We are very close

    • JasonDJ@lemmy.zip
      link
      fedilink
      English
      arrow-up
      16
      ·
      10 months ago

      Live is great but I don’t think it’d be feasible for most languages to be a real 1:1 translation in live.

      Even a 10s delay allows for the whole sentence/phrase to be captured and translated in entirety. A lot of languages can drastically change meaning due to a word on the other side of the sentence.

      • GreatAlbatrossA
        link
        fedilink
        English
        arrow-up
        7
        ·
        10 months ago

        The great thing about television, is that “live” is a flexible concept.
        The playback software could happily play 10 seconds ahead of what’s actually on the screen, and have plenty of time to translate like that.
        In the same way that we sometimes put delays into live events to allow the subtitling systems breathing room.

        • Trainguyrom@reddthat.com
          link
          fedilink
          English
          arrow-up
          4
          ·
          10 months ago

          In the same way that we sometimes put delays into live events to allow the subtitling systems breathing room.

          I’ve always heard this was because of the infamous Superbowl Janet Jackson wardrobe malfunction (where the malfunction was that only one nip was slipped and not both as was clearly intended)

      • azertyfun@sh.itjust.works
        link
        fedilink
        English
        arrow-up
        2
        ·
        edit-2
        10 months ago

        It’s already a thing with near-zero delay. MS Teams does it (dunno about the translation) and the QSMP Minecraft server has a bunch of livestreamers from different countries who use it for realtime translation.

        [EDIT: Live demo from today. Shit’s impressive.]

        What actually happens is that the current sentence gets “corrected” several times as you keep speaking. It’s a bit jittery and if the word order differs significantly then the translated sentence might be a bit wonky for a few seconds, and there are a few misses but overall it works really well; at least well enough that people who don’t speak each others’ language can have a conversation in their native tongues with essentially no more delay than reading speed. I can easily follow a livestream in a foreign language with the live subtitles (which was not the case a mere 6 months ago for any language other than English).

        • parody@lemmings.world
          link
          fedilink
          English
          arrow-up
          2
          ·
          2 months ago

          Amazing clip you posted seven months ago here. Doesn’t seem like it could even be any better now.

      • fatalError@lemmy.sdf.org
        link
        fedilink
        English
        arrow-up
        2
        arrow-down
        1
        ·
        10 months ago

        Live shouldn’t be used in a home setup anyway unless for something where interaction is required, like a teams call or twitch stream. Anything else can take a delay for the sake of preserving the meaning.

    • TwoCubed@feddit.de
      link
      fedilink
      English
      arrow-up
      8
      ·
      10 months ago

      I absolutely hate to watch subtitles appear word for word. So no, please no live captions.

      • RogueBanana@lemmy.zip
        link
        fedilink
        English
        arrow-up
        3
        ·
        10 months ago

        It doesn’t have to be live as in with the player but I imagine the audio could be loaded into the program simultaneously and have it produce cc for the entire movie as you watch it

    • Baku@aussie.zone
      link
      fedilink
      English
      arrow-up
      6
      ·
      10 months ago

      How does that work for people with non US/UK accents? I ask because all of the transcription software I’ve seen will work absolutely fantastically on even the most garbled and redneck American accents, and the vast majority of British ones too, but as soon as you get to Scottish/Welsh/German/Australian/really anywhere elses accents, it has a complete breakdown and you can’t make sense of it at all

    • pacoboyd@lemm.ee
      link
      fedilink
      English
      arrow-up
      4
      ·
      10 months ago

      Whisper AI is pretty darn good. I’ve used it to make subtitles for MST3K vids where nothing good exists and maybe only had to spend 10 minutes doing some clean up. It even recognizes when different people are speaking and breaks up the subs accordingly.

    • YoorWeb@lemmy.world
      link
      fedilink
      English
      arrow-up
      1
      ·
      10 months ago

      Imagine the next step though, soon AI will generate actors’ voices speaking in any language you want.

      • zygo_histo_morpheus@programming.dev
        link
        fedilink
        English
        arrow-up
        2
        ·
        10 months ago

        I don’t think I would use this actually, because I don’t see how an AI could capture the performance. I’m a sub over dub guy anyway, but at least someone making a dub has a sporting chance to make an interesting performance.