Geometry in Large Model Era

ECCV 2024 Workshop, Milano, Italy
September 30 PM, Brown 3

Geometric computing is crucial for many areas of application, concerning the extraction of topological and geometric information from shapes, scenes, and sequences, as geometric representations offer concise and intuitive abstractions, providing valuable methods for modeling, synthesis, compression, matching, and analysis. Recently, large foundation models have provided unprecedented opportunities for geometric computing and 3D vision, as shown in 2D to 3D content generation and open-vocabulary segmentation. We seek to bring together researchers from 3D computer vision, computational geometry, computer graphics, and deep foundation models to foster discussions on the latest breakthroughs, existing constraints, and future opportunities to develop a foundation model for geometric computing.

Our topics include but are not limited to:

  • Geometric computing techniques and their benefits to deep learning
  • Geometric losses and regularizations for model training and fine-tuning
  • Geometric feature learning on large-scale data
  • Geometry with open-vocabulary perception and interaction
  • Solving discrete and continuous geometric problems with large models
  • Foundation models for 3D content creation, perception, and geometric computing
  • Trends, opportunities, and future topics on geometric learning with large models

Keynote Speakers
Tom Funkhouser
Google
Marc Pollefeys
ETH & Microsoft
Andrea Vedaldi
Oxford
Angela Dai
TUM
Hao (Richard) Zhang
SFU
Qixing Huang
UT Austin
Schedule (tentative)
Welcome and Opening Remarks: Leonidas Guibas 14:00-14:10
Keynote Talk: Hao (Richard) Zhang 14:10-14:45
Keynote Talk: Marc Pollefeys 14:45-15:20
Keynote Talk: Qixing Huang 15:20-15:55
Coffee Break 15:55-16:10
Keynote Talk: Andrea Vedaldi 16:10-16:45
Keynote Talk: Angela Dai 16:45-17:20
Keynote Talk: Tom Funkhouser 17:20-17:50
Closing Remark 17:50-18:00
Organizers
Yichen Li
MIT
Congyue Deng
Stanford
Katie Luo
Cornell
Yonglong Tian
Google
Yue Wang
USC & NVIDIA
Jiajun Wu
Stanford
Minhyuk Sung
KAIST
Wojciech Matusik
MIT
Leonidas Guibas
Stanford
Sponsorship
Diamond $5000

  • Roll-up poster in the workshop room.
  • Acknowledgement at opening remark.
  • Acknowledgement and logo placement on the website

Platinum $3000

  • Acknowledgement at opening remark.
  • Acknowledgement and logo placement on the website

Gold $1500

  • Acknowledgement and logo placement on the website

The budget will be used for the following expenses:
  • Catered lunch for attendees and/or coffee & snacks for attendees
  • Registration for select (keynote) speakers or attendees
  • Swags for attendees
  • Support travel expenses
Any remaining budget from sponsorship will be rolled over to next year’s event.

Top