We received this query on our support desk email:
Do you have a self-generated subtitle tool that can be used for meet.coop meetings?
Looking forward to your answer,
This is a very interesting question. While Sophia didn’t mention it, the question could be referring also to simultaneous translated subtitles. That’s what others have requested in other moments. Let’s see what could be helpful. Inside each meeting room we can activate Closed Captions, which activates two elements: for moderators (translators will have to be promoted to moderator): 1) a realtime pad for each language where the translator can do their simultaneous translation. 2) all participants get the option to activate subtitles (closed captions) in the language of their choice (and in the color they wish).
Now what we’d need here is two things. First we’d need to have an Speech to Text conversion. That could be automatically outputting in the Closed Caption pad of the original language. And second would be an automatic translation service that outputs the translated text in the corresponding pad.
BTW I have seen announcements of BBB providers who have integrated these services, but in a proprietary way, not sharing the code with us. Question is how much effort it would be, and another, how many people would benefit from this.
Note that @leedscath mentioned some months ago how useful subtitles would be when holding assemblies in a noisy environment. That would only require part 1) of the solution. Part 2) would be valuable for multilingual environments, and much more economically viable I think, than hiring simultaneous translators for each language for each session. Then the automatic translations should be of sufficient quality. Maybe something to study in relation to our roadmap?
This general area - simultaneous interpretation and translation - is one I’d like to vote-up in our priorities. Apart from better matching the capabilities of some competing platforms, this would be a significant step in our intention to support multi-language exchange. As a multi-region infrastructure coop, this is important for the scope of our users, and thus for us.
I note the possibility of machine translation. I also note the multiple usefulnesses of text captions. But I also want to mention the elegant solution that has been adopted by MayFirst in their jitsi instance - a two-way language slider within the user interface, supporting live interpretation by humans.
So: a number of components to be scoped out, and ranked in a road map towards multi-language facilitation?
there’s a smaller UI issue with text captions . . they obscure a whole chunk of screen and thus, fo example, make it hard to see a presenter’s sides. A less space-hogging option could be helpful?
What about pointing them at a service like https://otter.ai in the first instance?
last year I was in zoom meetings supported with otter.ai. I have no clue though how one would integrate that right now (without some tech work) into meet.coop. AFAIK: otter.ai can use your microphone and perform a speech-to-text service. Then how does one get it synced with one’s meeting room for simultaneous subtitles?
as a quick test, I shared my mike with otter and started to speak. Here’s a sample so you get a feel for the quality. not perfect, but in many situations this might be good enough.
Hello, this evening and test so we can speak. And on my personal computers microphone when others shared was otter.ai is automatically transcribed to text. That’s wonderful. Now, what can I do with that text? Because I would ideally do synchronisation, simultaneous synchronisation with our meeting room and we’re not using Zoom Google meet now we’re having our own cooperatively organised. Big Blue Button service from meetup.com Up meetup.co op
Yeah, it’s almost there.
I’ve no idea how to integrate or even if it can be. I do know that people that I know use it to create a transcript from a recording. All I’m suggesting is that if people have a need in this area, this device may be worthwhile referring them to, as a matter of good customer service.