Machine Learning for Media Discovery Workshop

International Conference on Machine Learning (ICML)

July 18th, 2020 (PDT Timezone)

Messe Wien Exhibition Congress Center, Vienna, Austria

UPDATE: Due to COVID-19, ICML 2020 will be a virtual conference: A statement from the ICML organizers

ATTENDEES: ACCESS THE LIVESTREAM HERE

Invited Speakers

  • Jure Leskovec, Stanford University and Pinterest

  • Ed Chi, Google Brain

  • Eva Zangerle, Universität Innsbruck

  • Matthias Mauch, Apple Music

  • Delia Fano Yela, Chordify

Call For Papers

The ever-increasing size and accessibility of vast media libraries has created a demand more than ever for AI-based systems that are capable of organizing, recommending, and understanding such complex data.

While this topic has received only limited attention within the core machine learning community, it has been an area of intense focus within the applied communities such as the Recommender Systems (RecSys), Music Information Retrieval (MIR), and Computer Vision communities. At the same time, these domains have surfaced nebulous problem spaces and rich datasets that are of tremendous potential value to machine learning and the AI communities at large.

This year's Machine Learning for Media Discovery (ML4MD) aims to build upon the five previous Machine Learning for Music Discovery editions at ICML, broadening the topic area from music discovery to media discovery. The added topic diversity is aimed towards having a broader conversation with the machine learning community and to offer cross-pollination across the various media domains.

We invite the research community, from both industry and academia, to submit 2-page extended abstracts on topics such as:

  • Media (including Music, Movies, Podcasts, Series, etc.) recommendation and discovery

  • Media recommendation explainability at scale

  • Content-based and multimodal media recommender systems

  • Transfer learning and semi-supervised learning for media discovery

  • Fairness in recommendations

  • Bandits and reinforcement learning for media recommendations

  • Audio, video, image, and semantic content-based machine learning

  • Deep learning applications for computational audio and video research

  • Browsing and visualization of media datasets

  • Similarity metric learning

  • Learning to rank

  • Evaluation methodology

  • Modeling ambiguity and preference in media

  • Software frameworks and tools for deep learning in media

  • Automatic classification of media

  • AI-based media creation and machine creativity

  • Automatic media composition and improvisation

  • Media feature extraction

  • Pattern discovery

Template

Two page single-blind abstract submissions should be formatted according to the ICML template:

https://media.icml.cc/Conferences/ICML2020/Styles/icml2020_style.zip

Note that you will need to modify the template to use the accepted version of the template for authors to appear.

Only papers using the above template will be considered. Word templates will not be provided. Review is single-blind. A third page may be used for references only.

Submission Procedure

Papers should be submitted via EasyChair:

https://easychair.org/conferences/?conf=ml4md0

Poster Template

For posters we will follow the standard ICML requirements (24" x 36" in portrait orientation).

https://icml.cc/FAQ/PosterBoardSize

Contact

All questions about submissions should be emailed to Erik Schmidt (eschmidt@netflix.com) and/or Oriol Nieto (onieto@pandora.com).

Important Dates

  • Two-page abstracts deadline: May 1, 2020 June 5, 2020 June 12, 2020

  • Notification of Acceptance: May 9, 2020 June 22, 2020

  • Videos for oral presentation: June 30, 2020

  • Camera-Ready Deadline: July 1, 2020 July 12, 2020

  • Workshop Date: July 18, 2020

Presentation Videos

This year's event will be a livestream run by SlidesLive. Presenters that are selected for oral presentation will receive a link in order to record their presentation. The recording will capture both video of the presenter as well as the slides. We have a quick turnaround between acceptance and the video due date because SlidesLive requires time before the event for editing. During their recent work for the ICLR conference they received many poor quality and incomplete videos. We ask that presenters take extra care to consider camera position, lighting, and audio quality to make things easy on them. Documentation will be provided closer to the event.

Registration

Registration for ML4MD is done through ICML. This year ICML is charging a flat rate for access to all ICML events. The rate is $100 for participants or $25 for students.

Review Committee

  • Justin Basilico, Netflix

  • Ching-Wei Chen, Spotify

  • Sander Dieleman, Google DeepMind

  • James McInerney, Netflix

  • Filip Korzeniowski, Pandora

  • Mark Levy, Apple

  • Matt McCallum, Pandora

  • Matt McVicar, Apple

  • Vito Ostuni, Pandora

  • Justin Salamon, Adobe

Workshop Organizers

  • Erik M. Schmidt, Netflix

  • Oriol Nieto, Pandora

  • Fabien Gouyon, Pandora

  • Yves Raimond, Netflix

  • Katherine M. Kinnaird, Smith College

  • Gert Lanckriet, Amazon and UCSD