The first tutorial on
Neural Fields in Computer Vision


Full Day, June 20th, CVPR 2022 (New Orleans)


Slides are released!
Recording now available!


Abstract

Neural fields are emerging as a new signal representation for computer vision, computer graphics, and more. The research community on neural fields are ever more expanding, and there is a need to derive a taxonomy of the different components and techniques of neural fields to create a design space we can work within. In our tutorial, we will define the taxonomical basis of the neural fields design space, and do a deep dive into each of the components of neural fields. Our tutorial will also feature invited talks to showcase some of the unique applications of neural fields.

Schedule

Start Time End Time Chapter Section Presentor
9:00 AM 9:45 AM Introduction Introduction & Motivation Srinath / Towaki
Website & Moderators Or / Srinath
9:45 AM 10:15 AM Techniques Architectures Yiheng
10:15 AM 10:30 AM Break
10:30 AM 11:00 AM Techniques Hybrid Representations Towaki
11:00 AM 11:30 AM Forward Maps Yiheng
11:30 AM 12:00 PM Generalization Vincent
12:00 PM 12:30 PM Manipulation / Editing Or
12:30 PM 1:30 PM Lunch
1:30 PM 2:00 PM Applications 3D Reconstruction Vincent / James
2:00 PM 2:20 PM Generative Models Towaki / Shunsuke
2:20 PM 2:40 PM Humans Shunsuke
2:40 PM 3:00 PM Robotics Srinath
3:00 PM 3:10 PM Compression Towaki
3:15 PM 3:30 PM Break
3:30 PM 4:00 PM Invited Speakers Physics-Informed Neural Networks Zongyi Li
4:00 PM 4:30 PM Neural fields for Science Ellen Zhong
4:30 PM 5:00 PM Could Neural Fields be Useful for Content Creation? Alex Evans

Invited Speakers

Zongyi Li

Ellen Zhong

Alex Evans

Caltech
MIT
NVIDIA

Organizers

Yiheng Xie

Towaki Takikawa

Shunsuke Saito

Brown University / Unity
University of Toronto / NVIDIA
Meta Reality Labs

Or Litany

James Tompkin

Vincent Sitzmann

NVIDIA
Brown University
MIT

Srinath Sridhar

Brown University