[voiceinteraction] minutes August 25, 2021

https://www.w3.org/2021/08/25-voiceinteraction-minutes.html
(Also, note that Tobias was present -- I forgot to add him to the list.)
and below as text

   [1]W3C

      [1] https://www.w3.org/

                             - DRAFT -
                            voiceinteraction

25 August 2021

   [2]IRC log.

      [2] https://www.w3.org/2021/08/25-voiceinteraction-irc

Attendees

   Present
          Bev, Debbie, Jim, Jon, Tobias

   Regrets
          -

   Chair
          Debbie

   Scribe
          ddahl

Contents

    1. [3]publication publicity
    2. [4]meeting at TPAC
    3. [5]interfaces work

Meeting minutes

   bev: suggests a birds of a feather meeting at TPAC on running
   Community Groups

  publication publicity

   jon: paper has been publicized in OVON, and Bradley Metrock
   . Bret Kinsella

   debbie: included in standards column
   . for Speech Technology Magazine

   jim: should send note to Leonard Klie

   bev: could share on social media

  meeting at TPAC

   debbie: will have normal call during TPAC
   . only scheduled meeting is WebRTC, should we talk with them?

   debbie: it may be too low-level to concern us

   jim: would they be interested in what we're doing?
   . could it help them?

   debbie: could give them a use case

   jim: proposes introductory phone call

   debbie: no reason to wait for TPAC

   tobias: it's been integrated into a lot of software already,
   for example games

   debbie: books about MMI interaction and WebRTC

  interfaces work

   [6]https://w3c.github.io/voiceinteraction/
   voice%20interaction%20drafts/paArchitecture-1-2.htm#walkthrough

      [6] https://w3c.github.io/voiceinteraction/voice interaction drafts/paArchitecture-1-2.htm#walkthrough

   debbie: step 4 is between client and dialog manager

   <BC> Hello

   jim: will there be a path for control?
   . audio only, control, or tagged data?

   jim: question for Dirk -- what is included in item 4
   . also what about 5

   debbie: has to include all semantic information and enough
   information for the provider selection service to work
   . also forwards information to local ASR

   debbie: ask Dirk what is "received data" in step 5?
   . just the data from step 4/

   jim: lets talk about 5 next time, make a list of all data
   items, then decide if there's some structure

   debbie: what do we want to know about each data item?
   . type of media, source, target, confidence, input or output,
   timestamps

   <BC> What about language?

   jim: identifier for each data item

   <BC> Emotion Markup EML?

   debbie: information identifying the speaker, that raises
   security and privacy
   . could be embedded in EMMA or MMI Architecture

   debbie: what about step in the output path? how does that
   mapping happen?
   . step 14 is straightforward
   . could be streamed

   debbie: local processing is classic dialog system architecture
   . need to focus on steps 5, 9,10,11,12 and 13

   bev: let's not simplify 10 and 11

   debbie: could put together straw person table for data items

   jim: just a straw person proposal


   <BC> Thank you

Received on Wednesday, 25 August 2021 19:06:27 UTC