WebML CG Teleconference – 4 March 2021 - 15:00-16:00 UTC+0

➡️ https://github.com/webmachinelearning/meetings/blob/master/telcons/2021-03-04-agenda.md


WebML CG Teleconference – 4 March 2021 - 15:00-16:00 UTC+0
San Francisco (U.S.A. - California)     Thu, 4 March 2021       07:00   UTC-8 hours
Boston (U.S.A. - Massachusetts) Thu, 4 March 2021       10:00   UTC-5 hours
London (United Kingdom - England)       Thu, 4 March 2021       15:00   UTC+0 hours
Berlin (Germany)        Thu, 4 March 2021       16:00   UTC+1 hours
Helsinki (Finland)      Thu, 4 March 2021       17:00   UTC+2 hours
Shanghai (China)        Thu, 4 March 2021       23:00   UTC+8 hours
Tokyo (Japan)   Fri, 5 March 2021       00:00   UTC+9 hours
Corresponding UTC (GMT) Thu, 4 March 2021       15:00 UTC

Other locations: https://www.timeanddate.com/worldclock/fixedtime.html?iso=20210304T15


<https://github.com/webmachinelearning/meetings/blob/master/telcons/2021-03-04-agenda.md#logistics>Logistics

  *   Chair: Anssi
  *   Scribe: ?
  *   IRC: irc://irc.w3.org:6667/#webmachinelearning

  *   IRC web client: http://irc.w3.org/?channels=#webmachinelearning

  *   Call-in details: https://lists.w3.org/Archives/Member/internal-webmachinelearning/2020Apr/0000.html

  *   Minutes: https://www.w3.org/2021/03/04-webmachinelearning-minutes.html


<https://github.com/webmachinelearning/meetings/blob/master/telcons/2021-03-04-agenda.md#agenda>Agenda
<https://github.com/webmachinelearning/meetings/blob/master/telcons/2021-03-04-agenda.md#tag-review-feedback---open-prs>TAG review feedback - open PRs

NamedOutput mechanism clarification

  *   issue #140: https://github.com/webmachinelearning/webnn/issues/140

  *   PR #147: https://github.com/webmachinelearning/webnn/pull/147


Explainer feedback:

  *   issue #146: https://github.com/webmachinelearning/webnn/issues/146

  *   PR #148: https://github.com/webmachinelearning/webnn/pull/148


<https://github.com/webmachinelearning/meetings/blob/master/telcons/2021-03-04-agenda.md#tag-review-feedback---issues-with-active-discussion>TAG review feedback - issues with active discussion

Please take a look at these issues and make sure that you provide your feedback in comments. The editors will take your feedback into considerations when crafting PRs for these issues:

Isomorphic JS story, worker scope exposure?

  *   issue #142: https://github.com/webmachinelearning/webnn/issues/142


Prefix generic interface names?

  *   issue #141: https://github.com/webmachinelearning/webnn/issues/141


Ergonomics of the JS examples

  *   issue #139: https://github.com/webmachinelearning/webnn/issues/139


String enum for activations

  *   issue #138: https://github.com/webmachinelearning/webnn/issues/138


Clarify which view/reshape like function are expected to copy

  *   issue #137: https://github.com/webmachinelearning/webnn/issues/137


Constructor or builder pattern for model building?

  *   issue #136: https://github.com/webmachinelearning/webnn/issues/136


getNeuralNetworkContext() and createModelBuilder() params

  *   issue #135: https://github.com/webmachinelearning/webnn/issues/135


<https://github.com/webmachinelearning/meetings/blob/master/telcons/2021-03-04-agenda.md#tag-review-feedback---issues-still-looking-for-reactions>TAG review feedback - issues still looking for reactions

Please also check out these issues that have not received reactions yet:

https://github.com/webmachinelearning/webnn/issues?q=is%3Aissue+is%3Aopen+label%3Atag-tracker+comments%3A%3C1+no%3Aassignee


<https://github.com/webmachinelearning/meetings/blob/master/telcons/2021-03-04-agenda.md#privacy-and-security-review>Privacy and Security review

Please review our collective response to the Privacy IG feedback to date: https://github.com/webmachinelearning/webnn/issues/119#issuecomment-781508856


Based on this feedback, we should add a Security and Privacy Considerations to the spec, to be discussed: https://github.com/webmachinelearning/webnn/issues/122


<https://github.com/webmachinelearning/meetings/blob/master/telcons/2021-03-04-agenda.md#operation-specific-apis-proposal-next-steps>Operation-specific APIs proposal next steps

Discuss possible new features to WebNN API to satisfy the requirements of this proposal (@huningxin, @wchao1115):

  *   Direct access to the convolution op
  *   Support for native tensor types, GPU buffers
  *   Device preference setting
  *   Optimization for single-op graph execution

Please provide your comments: https://github.com/webmachinelearning/proposals/issues/2


<https://github.com/webmachinelearning/meetings/blob/master/telcons/2021-03-04-agenda.md#web-native-update>web-native update

An update on webnn-native open-source project progress and plans (@huningxin):

https://github.com/webmachinelearning/webnn-native