Data Centric AI / Workshop Nov 17-18 2021

The goal of this workshop is to bring together a new community of researchers, practitioners, organizations and individuals, and catalyze interest in the emerging discipline of Data-Centric AI.


This is a joint event between Stanford HAI and the ETH AI Center.

Stanford HAI        ETH AI Center

Recorded sessions

US Day 1

European Day

US Day 2

What is Data-Centric AI?

Creating the appropriate training and evaluation data is often the biggest challenge in developing AI in practice. This workshop will explore challenges and opportunities across the data-for-AI pipeline. We will discuss recent advances in curating, cleaning, annotating and evaluating datasets for AI. We will also investigate questions that arise from data regulations, privacy and ethics. The goal of the workshop is to help build an intellectual foundation for the emerging and critically important discipline of data-centric AI.

This event will be held virtually and is free to register.

Program Overview

Nov 17 2021 09:00 - 11:00 (Pacific Time)

US Day 1
The first session will be streamed from the US (pacific time) and will contain two keynote talks, a live discussion with both keynote speakers, as well as multiple student spotlight talks. Each talk will be followed by a online live Q&A session.

Nov 18 2021 13:00 - 18:00 (Central European Time)

European Day
The second part will be streamed from Europe (central european time) and will contain four keynote talks and multiple student spotlight talks, each talk will be followed by a online live Q&A session. This session will also include a live panel discussion along with a start-up session.

Nov 18 2021 09:00 - 11:00 (Pacific Time)

US Day 2
The third and final session will be streamed from the US (pacific time) and will contain two keynote talks, a fireside chat including both keynote speakers, as well as multiple student spotlight talks. Each talk will be followed by a online live Q&A session.

Keynote Speakers

Wed. Nov 17, US Day 1

9:00am - 9:15am (PT): Welcome and Introduction: What is Data-Centric AI?

James Zou

9:15am - 9:40am (PT): Keynote Talk

Matei Zaharia

9:40am - 10:10am (PT): Research Spotlights

10:10am - 10:35am (PT): Keynote Talk

Katharina Borchert

10:35am - 11:00am (PT): Keynote Talk

Curt Langlotz

11:00am - 11:05am (PT): Closing Remarks

James Zou

Thu. Nov 18, European Day

1:00pm - 1:05pm (CET): Welcome and Introduction

Ce Zhang

Session: Data Quality for AI

1:05pm - 1:30pm (CET): Keynote Talk

Felix Naumann

1:30pm - 2:00pm (CET): Research Spotlights

Cedric Renggli & Bojan Karlaš

Session: Data-Centric AI Research

2:00pm - 2:25pm (CET): Keynote Talk

Matthias Boehm

2:25pm - 2:45pm (CET): Research Spotlights

2:45pm - 3:00pm (CET): Keynote Talk

Debojyoti Dutta

Session: Start-up Session

3:00pm - 4:00pm (CET): Start-up session

Dicussion with the following startups: LIGHTLY, Syntheticus, Modulos, and LatticeFlow

Session: Data Systems for AI Governance

4:00pm - 4:25pm (CET): Keynote Talk

Sebastian Schelter

4:25pm - 4:35pm (CET): Research Spotlight

4:35pm - 5:00pm (CET): Keynote Talk

Martin Vechev

5:00pm - 5:55pm (CET): Panel on AI & Regulation

Panel from a legal aspect including regulators from the Canton of Zurich, legal experts, and technical experts on data security. Panelists: Shweta Shinde, Fabian Streiff, Stephanie Volz, Alexander Ilic. Panel moderator: Shalini Trefzer

5:55pm - 6:00pm (CET): Closing Remarks

Ce Zhang

Thu. Nov 18, US Day 2

9:00am - 9:05am (PT): Welcome and Introduction

James Zou

9:05am - 9:30am (PT): DCAI Benchmark

Ce Zhang

9:30am - 10:00am (PT): Research Spotlights

10:00am - 10:30am (PT): Fireside Chat: Building Benchmark Datasets

Fei-Fei Li

10:30am - 11:00am (PT): Keynote Talk

Pietro Perona

11:00am - 11:05am (PT): Closing Remarks

James Zou

European Day Panelists

Organizers

Acknowledgments

This workshop would not be possible without the generous support and devotion from Stanford HAI and ETH AI Center, including Vanessa Parli, Celia Clark, Alex Ilic, and others.