– DRAFT –
Accessible Platform Architectures Working Group Teleconference

exit

jasonjgw: Have found tables in wiki page useful

https://en.wikipedia.org/wiki/Comparison_of_video_container_formats

jasonjgw: Different container formats can contain transcripts and subtitles in packages

janina: Differences between transcripts and captions which are not well defined in specs

janina: Metadata needs to define whether captions and transcripts are part of the package

jasonjgw: Need to review container formats more thoroughly.

Raja: One element is the intellectual property considerations for some of these container technologies

janina: Example of various IP licensing of content which limits descriptions from one vendor to another

janina: Where and why something is blocked could be contained in the metadata

Raja: Distinction between auto captions and human captions

Raja: For example, YouTube includes autocaptions

Raja: We need to confirm

Janina: Can a YouTube user turn on autocaptions?

Raja: Not sure if these are always provided, but the platform makes it possible to provide autocaptions

janina: Agree that we need to research and document these capabilities

jasonjgw: We can add more to our wiki page as we do our research

jasonjgw: We need to keep investigating over the next week or so...bring back next week

janina: Should also address in upcoming iterations of the EPUB standard

Media synchronization use cases and user needs.

Task Force publications relevant to metaverse standardization, and related issues.

jasonjgw: Not much to update since last time other than an additional reference that was added

janina: Do we also want to reference the MAUR?

jasonjgw: Would lean toward including it

janina: Content Useable may be generally applicable

jasonjgw: Further thoughts?

Media synchronization use cases and user needs.

Miscellaneous updates and topics.

jasonjgw: Other items we need to address?

janina: Discussion around plans to register accessible metadata for human readable

jasonjgw: We could help with taxonomy at some point

jasonjgw: Concerning IP issues around alternate media versions, there is a recent book on this

janina: Example of the DCMA exemption for accessibility

jasonjgw: Will send out the link to the reference

janina: Legal history on Fair Use statute which cites accessibility many years ago

jasonjgw: A lot happening on machine learning

jasonjgw: Recent research on emergent learning of AI, based on the theory of mind

Raja: Have noticed that there is less common data and a high chance of AI making up facts on low-incidence populations

Raja: There is a high chance of error in tools like ChatGPT

janina: Jutta Treviranus is a good resource on this

janina: She did a good presentation on AXE-CON about this

jasonjgw: When AI starts making attributions about intentions for individuals, a lot can go wrong

jasonjgw: On the other hand, AI is doing a better job on image descriptions

Raja: AI as a captioning assistant, where a human corrects any inaccuracies as it is produced, is also a useful direction

<janina> Jutta's presentation at: https://www.deque.com/axe-con/sessions/preparing-accessibility-for-the-future-and-the-present/

jasonjgw: Anything else??