PowerPoint and Skype to Get Live Captions, Subtitles

Posted on December 3, 2018 by Mehedi Hassan in Microsoft, Office 365, Skype with 6 Comments

Microsoft is introducing a major feature for PowerPoint and Skype. The company is launching live captioning and subtitles for both Skype and PowerPoint today.

On Skype, you will be able to enable live captions during calls. Skype will then automatically display captions in your call, and they will automatically scroll as people talk in your call. The app will let you scroll through all the captions on its own side window in the coming months.

On PowerPoint, the feature works similarly: as the presenter talks, you can display live captions underneath the slides. The feature apparently uses AI to automatically adapt to different content and context to improve the captions. Users will also be able to easily customize the size, position and appearance of the subtitles.

The new feature are supposed to improve inclusivity on both Skype and PowerPoint, so it will likely be a welcome addition for many. The feature will be available to Office 365 subscribers on PowerPoint in late January 2019, and it’s available on Skype today. Skype will also be getting support for translation of 20 languages in the coming weeks.

Tagged with , ,

Join the discussion!

BECOME A THURROTT MEMBER:

Don't have a login but want to join the conversation? Become a Thurrott Premium or Basic User to participate

Register
Comments (6)

6 responses to “PowerPoint and Skype to Get Live Captions, Subtitles”

  1. chump2010

    Just to ask a stupid question, but does this mean that Skype knows everything you say when you say something? I mean if they can transcribe your voice to captions, then surely that means they know everything you say. Not just what you type, but what you say too. Which is kind of scarey really.

    3 years ago
    0

    • MikeGalos

      In reply to chump2010:

      They certainly can do that and have been able to for years on ANY voice system.


      When I worked in the Speech team and on the Messenger team at Microsoft we consciously decided to NOT implement that to protect customer privacy but we did assume that Google had at least to the point of doing speech2text, scanning for keywords and adding them to the profiles of the users on the call. And that was a decade ago.

      3 years ago
      0

    • SvenJ

      In reply to chump2010: Well I guess it depends on what you mean by knowing everything you say. Of course it needs to listen to be able to produce the speech to text. While we have been doing speech to text for a long time, on machine, with things like Dragon Naturally Speaking, doing this with multiple separeated speakers, probably needs more power than most people have in the devices they'd want to use.
      Now, does MS (or Google, or Amazon) save every scrap of audio and processed text? Probably not. Does the backend 'know' what you said, or is it just 'translating it? Alexa routinely translates what I said perfectly, but still has no clue what I wanted. I'm sure there is some benefit to saving enough to do some analysis for improving recognition. In any case, it is your option to turn it on or not. If you are going to have a conversation with someone who is hearing impaired or in an environment that audio isn't practical, this could be super. You weigh the risk.


      3 years ago
      0

    • wright_is

      In reply to chump2010:

      With the typing, it can be end-to-end, but for captions it will have to channel the audio (at least) through their cloud AI, so it is no longer end-to-end.

      3 years ago
      0

    • FalseAgent

      In reply to chump2010:

      weird how no one asks this of every single youtube video that has automatic transcribing using Google's AI

      3 years ago
      0

  2. chump2010

    Thank you to all three for your answers. I knew about the email scanning that every provider did. I knew that the various personal assistants were obviously recording your voice etc. I did not realise that even in Skype, everything potentially is being parsed by an AI of some kind - even when it is voice. I assumed that perhaps they were reading my conversations for advertising purposes (though that is not made clear). It is just the idea that they also know everything that is said as well potentially recording all my voice interactions too, is a bit scarey. At least for me. Many thanks for your thoughts :)

    3 years ago
    0

Leave a Reply

We use cookies to improve your browsing experience.

Find out more about our cookie policy here.