Meeting minutes
TPAC 2023 meeting plans.
janina: We have provided some good input to APA on groups to meet with
janina: Some thinking about which groups may have major releases coming
janina: WCAG is heading toward a major transition toward 3.0
janina: We have the well known URI and the ADAPT work
janina: The pronunciation work is another one
janina: EPUB is another group that has some big possibilities opening up
janina: These are good opportunities for TPAC meetings where the future possibilities can be discussed
janina: There may also be things like synchronization of musical scores, syncing or multiple language captions is another
<janina> Proposed Charter for the Media Interest Group: https://
Incorrect link?
<janina> Correct URI is: https://
jasonjgw: We will be working on TPAC scheduling for these groups, and hoping we can all attend as we are able
janina: We'll be proving the form for meeting planning on Friday morning, so if anyone has additional thoughts send that to the list
Collaboration Tools Accessibility User Requirements.
jasonjgw: Still looking forward to upcoming COGA comments
scott_h: Still working on some things on my end
Structure and synchronization of media resources.
jasonjgw: Meeting with Timed Text is planned for TPAC
janina: Interlinear tracking of content has not thus been considered, but we may want to broach that topic as well
jasonjgw: We should be able to add that to the wiki
janina: To Raja; what about having more that one language in the captions?
janina: We are thinking about multiple language tracks in a synchronized manner
janina: Would that parallel work in sign language as well?
Raja: Line by line might be possible - some players can perhaps do this. Would this be prerecorded with time stamps?
janina: Yes
Raja: That sounds use. I would support this at the sentence level.
jasonjgw: I will update the wiki with that as well.
Raja: two streaming video lines as opposed to two lines of text?
janina: Yes, because word order won't be the same in different languages.
Raja: so we could have two signing videos?
janina: That would be the concept...multiple video blocks
raja: Video 1 would be the person speaking and video 2 would be the signer, and time stamps would keep these synchronized.
Raja: It would be tricky because multiple time stamps would need to be tracked
janina: Perhaps once Jason has something in the Wiki, he can post to list and Raja can reply
Raja: Multiple signing video streams with text captions all with timestamps to link for synchronization
jasonjgw: Will give this an attemp
janina: The main issue will be uptake by user agents
Universal Design for Learning Guidelines: request from WAI-Adapt Task Force.
jasonjgw: They are still looking at questions, so not much to say
janina: Not much to say now
Natural language interface accessibility User Requirements.
jasonjgw: Some of the underlying research will need work
janina: We could perhaps ping Josh about this
jasonjgw: There are a lot of comments that have been generated, but responses will require a good bit of time
jasonjgw: Will need to keep bringing this back
Miscellaneous updates and topics.
janina: Still waiting for something from the FAST taskforce, but nothing at this point
jasonjgw: Some news items around more efficient automated TTS tools
janina: Ongoing discussion about BLIS symbols being added to Unicode
scott_h: Recent conversations have included whether there should be an "accessibility of AI" guidelines and so forth?
scott_h: But maybe this fall into other spaces?
janina: Perhaps this fits into Natural Language
jasonjgw: There is some meshing which would be in scope for natural language
scott_h: But some connections are image based...not language.
janina: It may be that adding "AI" into the title might help stimulate more discussion
janina: It seems like there will always be a command in language
janina: While the output may not be text, the command will have to be language based for the AI to understand what it is supposed to do
jasonjgw: We may want to consider a title and scope change
jasonjgw: This may also connect to XR, so that's another angle