index.html

---
title: Workshop overview
layout: subpage
---
    <main id="main" class="main">
      <section id="home">
        <section id="intro">
          <h2>Introduction</h2>
          <p>Over the months of August and September 2020, W3C conducted a virtual <a href="https://www.w3.org/2003/08/Workshops/">workshop</a> to pave the way for <strong>enriching the Open Web Platform with better
              foundations for machine learning</strong>.</p>

          <h2>Outcomes &amp; Proceedings</h2>
          <p><a href="presentations.html">34 presentations were recorded and published</a>.</p>
          <p>These presentations triggered <a href="https://github.com/w3c/machine-learning-workshop/issues?q=is%3Aissue+label%3A%22Discussion+topic%22+">25 online workstreams</a> which were discussed in <a href="proceeding.html">4 minuted live sessions</a>.</p>
          <p>The <strong><a href="report.html">Workshop Report</a></strong> summarizes the key findings from the workshop.</p>
          
          <h2 id="goals">
            What was the purpose of this workshop?
          </h2>
          <p>
            The primary goal of the workshop was to bring together providers of
            machine learning toolkits and framework providers with Web platform
            practitioners to <strong>enrich the Open Web Platform with better
            foundations for machine learning</strong>.
          </p>
          <p>
            The secondary goals of the workshop were as follows:
          </p>
          <ul id="secondary-goals">
            <li>Understand how machine learning fits into the Web technology
            stack,
            </li>
            <li>Understand how browser-based machine learning fits into the
            machine learning ecosystem,
            </li>
            <li>Explore the impact of machine learning technologies on Web
            browsers and Web applications,
            </li>
            <li>Evaluate the opportunities for standardization around machine
            learning APIs and formats.
            </li>
          </ul>
          <h2 id="topics">
            🧐 What topics were in scope?
          </h2>
          <p>
            The following topics were identified as in scope
          </p>
          <ul id="proposed-topics">
            <li>
              Opportunities and Challenges of Browser Based Machine Learning
              <ul>
                <li>Privacy-First approach to machine learning
                </li>
                <li>Real-time in-browser Machine Learning
                </li>
                <li>Performance, compatibility, JS environment gaps
                </li>
                <li>Domain-specific compilers for Machine Learning
                </li>
              </ul>
            </li>
            <li>
              Web
              Platform Foundations for Machine Learning
              <ul>
                <li>Web Platform: a 30,000 foot view
                </li>
                <li>Web Platform and JS environment constraints
                </li>
                <li>Bringing Machine Learning to the JS ecosystem with Machine
                Learning libraries
                </li>
                <li>Accelerated graphics and compute APIs for Machine Learning
                </li>
                <li>Fast, portable code with WebAssembly / WASI-nn
                </li>
                <li>Access purpose-built Machine Learning hardware with WebNN
                </li>
              </ul>
            </li>
            <li>
              Machine
              Learning Experiences on the Web: A Developer's Perspective
              <ul>
                <li>On-device training in browser
                </li>
                <li>Datasets on the Web & Schema.org vocabularies
                </li>
                <li>Interoperability of Machine Learning models for the Web
                </li>
                <li>High-level load & run model vs low-level graph builder API
                </li>
                <li>Integration of models and in-browser data sources sensors,
                AV
                </li>
                <li>Considerations when deploying models to the web
                </li>
                <li>TensorFlow.js
                </li>
                <li>ONNX.js
                </li>
                <li>Magenta.js
                </li>
                <li>ML5.js
                </li>
                <li>Paddle.js
                </li>
                <li>Machine Learning in Web Architecture
                </li>
              </ul>
            </li>
            <li>
              Machine
              Learning Experiences on the Web: A User's Perspective
              <ul>
                <li>Teachable Machine & Project Euphonia
                </li>
                <li>Visualization of deep networks, "human-interpretable neural
                nets"
                </li>
                <li>Web a11y opportunity
                </li>
                <li>Cross-industry case studies
                </li>
                <li>Media technologies roadmap for the Web
                </li>
                <li>Enhancing media experiences with Machine Learning
                </li>
                <li>Making art with Machine Learning
                </li>
                <li>Making music with Machine Learning
                </li>
                <li>Teaching machines how people speak
                </li>
              </ul>
            </li>
            <li>
              <span class="type type-plenary">plenary</span> Machine Learning
              Consensus Landscape
              <ul>
                <li>Who is doing what: what's happening in standards, what's
                happening in related open source projects.
                </li>
              </ul>
            </li>
          </ul>
        </section>
        <section>
          <h2 id="w3c">
            🌐 What is W3C?
          </h2>
          <p>
            W3C is a voluntary standards consortium that convenes companies and
            communities to help structure productive discussions around
            existing and emerging technologies, and offers a Royalty-Free
            patent framework for Web Recommendations. We focus primarily on
            client-side (browser) technologies, and also have a mature history
            of vocabulary (or “ontology”) development. W3C develops work based
            on the priorities of our members and our community.
          </p>
        </section>
        <section>
          <h2 id="program">
            👋 Program Committee
          </h2>
          <h4>
            Chairs
          </h4>
          <ul class="pc">
            <li>Kelly Davis (Mozilla)
            </li>
            <li>Anssi Kostiainen (Intel)
            </li>
          </ul>
          <h4>
            Committee
          </h4>
          <ul class="pc">
            <li>Göran Eriksson (Ericsson)
            </li>
            <li>Dominique Hazaël-Massieux (W3C)
            </li>
            <li>Ningxin Hu (Intel)
            </li>
            <li>Dean Jackson (Apple)
            </li>
            <li>Sangwhan Moon
            </li>
            <li>Roy Ran (W3C)
            </li>
            <li>Georg Rehm (DFKI)
            </li>
            <li>Amy Siu (Beuth University of Applied Sciences, Berlin)
            </li>
            <li>Nikhil Thorat (Google)
            </li>
          </ul>
        </section>
      </section>
      <section>
        <h3>
          Sponsors
        </h3>
        <p><a href="https://www.futurice.com/"><img src="images/futurice.png" alt="futurice" width=400></a></p>
      </section>
    </main>