As part of our meeting we will have a focused process of working in small groups in order to make tangible progress in advancing computational and developmental approaches to learning. Each working group will be meeting throughout the week, and your output will be structured in the form of two presentations:
10/19 Wed morning: Initial report-back, each group shares the research topic they have selected, their goal with respect to that topic, and an overview in the form of 1-2 slides
10/21 Fri morning: Final report-back, each group presents on the output of their activities for the week to the rest of the attendees.
The topics that you voted for were framed in broad terms to give each group plenty of scope to identify a specific topic of interest. There are many possibilities for the output of each group. For example, you may want to define a specific research problem of interest and make progress in framing it and identifying lines of attack. Or you may want to identify one or more open research problems in a particular area and build the case for their importance, and so forth. Each group is free to define its area of focus and its goals to align with their interests. It may be that some of these group activities will lead to white papers or other publishable documents, and we will support these activities wherever there is interest. We will also be utilizing the outputs of your working group’s activities as part of the meeting report (a requirement for Dagstuhl).
Working Group 1
What are the connections between current computational research in self-supervised, weakly-supervised, and continual machine learning, and analogous developmental learning processes in humans and animals? (2 groups 1.1 and 1.2)
Working Group 2
What is the role of computational models of learning (e.g., object recognition, machine perception, and reinforcement learning) in advancing developmental science? Can computational tools enable new developmental research questions? What kinds of data should developmental scientists produce that would be the most useful for computational approaches? (2 groups 2.1 and 2.2)
Working Group 3
What is the role of multimodal learning (learning from diverse signal types such as visual, audio, touch, force, etc.) in development? What are the challenges and opportunities in multimodal machine learning?
Uri Hasson
Rhodri Cusack
Celeste Kidd
Marc'Aurelio Ranzato
Stefan Stojanov
Anh Thai
Daniel Swingley
Chen Yu
Anne Warlaumont
Jim Rehg
Emmanuel Dupoux
Gert Westermann
Rebecca Saxe
Michael Frank
Maureen de Seyssel
Maithilee Kunda
Ingmar Visser
Marvin Lavechin
Pierre-Yves Oudeyer
Hana D'Souza
David Crandall
Abdellah Fourtassi
Alejandrina Cristia
Olivier Sigaud
Clément Romac
Atsushi Nakazawa
Judy Hoffman
Casey Lew-Williams
Felix Hill
Jelena Sucevic
Eon-Suk Ko
Hiromichi Hagihara
Thomas Carta