Deep Learning and Design

Saturday 2nd July 2022, 2:00 pm to 5:30 pm

Workshop Chairs

Jianxi Luo | jianxi_luo@sutd.edu.sg

Immanuel Koh | immanuel_koh@sutd.edu.sg

Workshop Committee

Sean Hanna, University College London

Pieter Pauwels, Eindhoven University of Technology

Christopher McComb, Carnegie Mellon University

Mark Fuge, University of Maryland

Aim and Content

The rapid advances in deep learning in the past decade have provided many new opportunities for design research and practice. The goals of this workshop are to elucidate the current state of developing and adopting deep learning techniques in design research and practice, identify the gaps, challenges, and opportunities, and formulate a research framework and research agenda for deep learning in design toward the future. The workshop also aims to bring together researchers from different design disciplines who share the interests in design + deep learning and can potentially collaborate. The workshop might also result in a white paper, which the workshop participants will be invited to contribute to.

Questions relevant to this workshop include but are not limited to:

  • What is the current state of design research that uses deep learning techniques?

  • Which deep learning techniques, with what data sources, have been used or can be potentially used, to support what design tasks (i.e., design representation, design generation, design process management), at which design process phases (e.g., conceptual design, detailed design), in which design fields (e.g., architecture, engineering)?

  • How should human and deep learning-based AI interact in the design process?

  • What are the near-term opportunity and long-term directions for design research with deep learning?

Participation

The workshop is open to any design researchers with an interest in deep learning. Prior knowledge and experience in deep learning will be valuable for the discussions but not required. Participants may come prepared to briefly describe (1 minute) their prior research or research ideas on deep learning for-and-in design.

Workshop Format

  1. Brief introductions and sharing from all workshop attendees (30 minutes)

  2. Lightning talks (30 minutes)

  3. Breakout group discussions of all participants (60 minutes)

  4. Preparation of group presentations (15 minutes)

  5. Group presentations (30 minutes)

  6. Collating workshop findings (10 mins)

  7. Closing (5 minutes)