Visual Pre-Training for Robotics

CVPR 2023 Workshop
Day: June 19, 2023
Time: 9:15am - 5:30pm
Location: West 220-222

The great vision scientist, James J. Gibson, famously said “We see in order to move and we move in order to see.” But can we learn to see before we learn to move? And how far can we move if we first learn to see?

We are excited to have a lineup of speakers from computer vision, machine learning, and robotics. We hope that this workshop will help attract the broader CVPR community to this important and exciting topic.



Time Speaker Talk Title
9:15 - 9:30 am Introduction  
9:30 - 10:00 am Kristen Grauman From Goals to Grasps:
Learning About Action from People in Video
10:00 - 10:30 am Jitendra Malik Visual Representations for Robotics
10:30 - 11:00 am Morning Coffee  
11:00 - 11:30 am Dieter Fox Toward Foundational Manipulation Skills
11:30 - 12:00 pm Josef Sivic Learning Manipulation Skills from Instructional Videos
12:00 - 2:00 pm Lunch  
2:00 - 2:30 pm Chelsea Finn How to Generalize Your Robot with Diverse Offline Data
2:30 - 3:00 pm David Fouhey From Hands in Action to Possibilities of Interaction
3:00 - 3:30 pm Afternoon Coffee  
3:30 - 4:00 pm Abhinav Gupta Watch, Learn, Manipulate:
Learning Manipulation in the Wild from Passive Videos
4:00 - 4:30 pm Alexander Kirillov Segment Anything
4:30 - 5:00 pm Sergey Levine Robotic Learning with Large Datasets:
Towards General-Purpose Models for All Robots
5:00 - 5:30 pm Discussion Panel