From ee2f16b88b26960c3deeb3f1a4bc0b7083c82661 Mon Sep 17 00:00:00 2001 From: Dominique Hazael-Massieux Date: Fri, 3 Feb 2023 11:13:44 +0100 Subject: [PATCH 01/16] Fix links to WG pages --- charter.html | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) diff --git a/charter.html b/charter.html index 51de72c..d3ecc83 100644 --- a/charter.html +++ b/charter.html @@ -103,13 +103,13 @@

"https://github.com/w3c/machine-learning-charter/issues">issues.

- The mission of the Web Machine Learning + The mission of the Web Machine Learning Working Group is to develop APIs for enabling efficient machine learning inference in the browser.

@@ -259,7 +259,7 @@

To avoid overlap with existing work, alignment with the Basic Linear Algebra Subprograms (BLAS) interface is out of scope. The WebGPU shaders and WebAssembly SIMD are expected to address the - BLAS compatibility requirement, see the Coordination section for + BLAS compatibility requirement, see the Coordination section for details.

@@ -571,7 +571,7 @@

Information about the group (including details about deliverables, issues, actions, status, participants, and meetings) will be - available from the Web Machine Learning Working Group home + available from the Web Machine Learning Working Group home page.

@@ -643,8 +643,7 @@

standards, W3C seeks to issue Recommendations that can be implemented, according to this policy, on a Royalty-Free basis. For more information about disclosure obligations for this group, please - see the W3C Patent - Policy Implementation. + see the licensing information.

From 72cd9b7e100f09b566d8a98477e4a6c3a3f9ed01 Mon Sep 17 00:00:00 2001 From: Dominique Hazael-Massieux Date: Fri, 3 Feb 2023 11:14:27 +0100 Subject: [PATCH 02/16] Tweak background and motivation incl highlighting RTC as a use case --- charter.html | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/charter.html b/charter.html index d3ecc83..230b377 100644 --- a/charter.html +++ b/charter.html @@ -187,10 +187,11 @@

Motivation and Background

the browser (as opposed e.g. to in the cloud) enhances privacy, since input data such as locally sourced images or video streams stay within the browser's sandbox. Local processing also enables machine learning use - cases that require low latency, such as object detection in immersive - web experiences. + cases that require low latency, such as object detection in real-timee + communications and immersive web experiences.

- Currently, machine learning inference in the browser uses the WebGL + Currently, machine learning inference in the browser uses WebAssembly + and the WebGL graphics API with limited or no access to platform capabilities beneficial for ML such as CPU parallelism, general-purpose GPU, or dedicated ML hardware accelerators. @@ -217,7 +218,7 @@

for hardware execution;
  • Allows to setup input from various sources on the Web, e.g. array - buffers, media streams, schedule the asynchronous hardware execution, + buffers and media streams, schedule the asynchronous hardware execution, and retrieve the output when hardware execution completes.
  • From 1d6c31ff79cd4464d28cf0b9e9905b1dba33647d Mon Sep 17 00:00:00 2001 From: Dominique Hazael-Massieux Date: Fri, 3 Feb 2023 11:14:52 +0100 Subject: [PATCH 03/16] Use present to describe scope --- charter.html | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/charter.html b/charter.html index 230b377..ce3492b 100644 --- a/charter.html +++ b/charter.html @@ -230,8 +230,8 @@

    Recognition.

    - The APIs in scope of this group will not be tied to any particular - platform and will be implementable on top of existing major platform + The APIs in scope of this group are not be tied to any particular + platform and are to be implementable on top of existing major platform APIs, such as Android Neural Networks API, Windows DirectML, and macOS/iOS Metal Performance Shaders and Basic Neural Network Subroutines. From 15070a0bb4887d0818da1d3dc78478b92853aa32 Mon Sep 17 00:00:00 2001 From: Dominique Hazael-Massieux Date: Fri, 3 Feb 2023 11:15:10 +0100 Subject: [PATCH 04/16] Clarify what WebGL interop is out of scope --- charter.html | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/charter.html b/charter.html index ce3492b..0c13948 100644 --- a/charter.html +++ b/charter.html @@ -264,9 +264,8 @@

    details.

    - Interoperability between the WebNN and WebGL APIs is out of scope. + Integration between the WebNN and WebGL APIs is out of scope, as the group focuses its efforts on integration with WebGPU APIs.

    -

      From 191a1c44c9c48898754cf9888c88e860850d3874 Mon Sep 17 00:00:00 2001 From: Dominique Hazael-Massieux Date: Fri, 3 Feb 2023 11:15:32 +0100 Subject: [PATCH 05/16] Document expected data on existing draft --- charter.html | 22 +++++++++------------- 1 file changed, 9 insertions(+), 13 deletions(-) diff --git a/charter.html b/charter.html index 0c13948..1bb2ca4 100644 --- a/charter.html +++ b/charter.html @@ -295,24 +295,20 @@

      inference that can take advantage of hardware acceleration.

      - Draft state: Working Draft + Draft state: W3C Working Draft

      -

      - Adopted Draft: The title, stable URL, - and publication date of the - Adopted Draft which will serve as the basis for - work on the deliverable. +

      + Expected completion: Q1 2025

      - Exclusion Draft: The title, stable - URL, and publication date of the most recent - Exclusion Draft. + Adopted Draft: https://www.w3.org/TR/2023/WD-webnn-20230124/, 2023-01-24 (this will change before the AC review starts)

      -

      - Expected completion: Q1 2025 +

      + Exclusion Draft: https://www.w3.org/TR/2021/WD-webnn-20210622/">, 2021-06-22
      + Exclusion period began 2021-06-22; Exclusion period ended 2021-11-19.

      +

      Exclusion Draft Charter: https://www.w3.org/2021/04/web-machine-learning-charter.html +

      From 6c521051ff775fd63af60c2f7e99d9767bef9280 Mon Sep 17 00:00:00 2001 From: Dominique Hazael-Massieux Date: Fri, 3 Feb 2023 11:16:10 +0100 Subject: [PATCH 06/16] Re-add a timeline section it remains expected and helps with reviews --- charter.html | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/charter.html b/charter.html index 1bb2ca4..94d3e34 100644 --- a/charter.html +++ b/charter.html @@ -373,6 +373,13 @@

      +
      +

      Timeline

      +
        +
      • Q2 2023: Candidate Recommendation for WebNN
      • +
      • Q1 2025: Recommendation for WebNN
      • +
      +

      From 8a7230a591d31ab4dbcfbb97075f1bda55f68a61 Mon Sep 17 00:00:00 2001 From: Dominique Hazael-Massieux Date: Fri, 3 Feb 2023 11:16:51 +0100 Subject: [PATCH 07/16] Align testing considerations with latest charter template --- charter.html | 9 ++------- 1 file changed, 2 insertions(+), 7 deletions(-) diff --git a/charter.html b/charter.html index 94d3e34..d7dd455 100644 --- a/charter.html +++ b/charter.html @@ -392,7 +392,7 @@

      specification is expected to have at least two independent implementations of every feature defined in - the specification. + the specification, where interoperability can be verified by passing open test suites, and two or more implementations interoperating with each other. In order to advance to Proposed Recommendation, each normative specification must have an open test suite of every feature defined in the specification..

      Each specification should contain a section detailing all known @@ -404,7 +404,7 @@

      There should be testing plans for each specification, starting from - the earliest drafts. + the earliest drafts. To promote interoperability, all changes made to specifications in Candidate Recommendation or to features that have deployed implementations should have tests. Testing efforts should be conducted via the Web Platform Tests project.

      Each specification should contain a section on accessibility that @@ -412,11 +412,6 @@

      features can be used to address them, and recommendations for maximising accessibility in implementations.

      -

      - To promote interoperability, all changes made to specifications - should have tests. -

      From a4a9311a4ff62e214d4ef7b25486341261442659 Mon Sep 17 00:00:00 2001 From: Dominique Hazael-Massieux Date: Fri, 3 Feb 2023 11:17:27 +0100 Subject: [PATCH 08/16] Clarify coordination with WebGPU WG --- charter.html | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/charter.html b/charter.html index d7dd455..3556c3b 100644 --- a/charter.html +++ b/charter.html @@ -468,11 +468,11 @@

      Group
      - The GPU for the Web Working Group defines a WebGPU Shading - Language that may be used to implement traditional machine + The GPU for the Web Working Group defines access to GPU devices with the WebGPU API, and a that may be used to implement traditional machine learning algorithms efficiently. The Web Machine Learning Working - Group should coordinate with this group to avoid overlap and - to enable interoperability between the WebNN and WebGPU APIs. + Group will coordinate with this group to avoid overlap and + to ensure proper integration of WebNN with WebGPU APIs.
      @@ -484,7 +484,7 @@

      The WebAssembly Community Group incubates a proposal for a 128-bit SIMD support in WebAssembly that can be used to implement traditional machine learning algorithms efficiently. The Web - Machine Learning Working Group should coordinate with this group + Machine Learning Working Group will coordinate with this group to avoid overlap.
      WebRTC Working Group
      From 740b1ab9649541a0d524c7458cf67b1e0a42a530 Mon Sep 17 00:00:00 2001 From: Dominique Hazael-Massieux Date: Fri, 3 Feb 2023 11:17:46 +0100 Subject: [PATCH 09/16] Align with charter template --- charter.html | 16 ++++++---------- 1 file changed, 6 insertions(+), 10 deletions(-) diff --git a/charter.html b/charter.html index 3556c3b..143f29a 100644 --- a/charter.html +++ b/charter.html @@ -709,6 +709,10 @@

      +
      +

      Change log

      +

      Changes to this document are documented in this section.

      +

      @@ -717,16 +721,8 @@

      Dominique Hazael-Massieux

      Motivation and Background

      the browser (as opposed e.g. to in the cloud) enhances privacy, since input data such as locally sourced images or video streams stay within the browser's sandbox. Local processing also enables machine learning use - cases that require low latency, such as object detection in real-timee + cases that require low latency, such as object detection in real-time communications and immersive web experiences.

      Currently, machine learning inference in the browser uses WebAssembly From b537d2ce703d5e15522e386750d6a3073721b8f0 Mon Sep 17 00:00:00 2001 From: Dominique Hazael-Massieux Date: Fri, 3 Feb 2023 11:48:53 +0100 Subject: [PATCH 11/16] Align with more changes to charter template --- charter.html | 39 ++++++++++++++++++--------------------- 1 file changed, 18 insertions(+), 21 deletions(-) diff --git a/charter.html b/charter.html index ac583fb..f4225fa 100644 --- a/charter.html +++ b/charter.html @@ -55,6 +55,7 @@