XFeiF / ComputerVision_PaperNotes

📚 Paper Notes (Computer vision)
1 stars 0 forks source link

18CVPR| Unsupervised Visual Learning Tutorial #12

Closed XFeiF closed 3 years ago

XFeiF commented 4 years ago

[part1] [part2]

XFeiF commented 4 years ago

slides

XFeiF commented 4 years ago

Outline

Part 1: Introduction: Unsupervised learning from vision
Part 2: Representation learning from images and videos
Part 3: Unsupervised learning from 3D
Part 4: Self-supervised learning for Robotics

XFeiF commented 4 years ago

Part 1 Highlights

Unsupervised visual learning usefulness.

Learning from videos (sequentially correlated data) or images (spatially correlated data) provides a lot of information for feature representation!

Learning from 3D - very valuable as it provides structure (and constraints) for learning.

Learning with self-supervision and interactions - provides more task-specific "supervision"; can directly learning outcomes; provides better features for subsequent tasks.

image

XFeiF commented 4 years ago

Part 2 Highlights

Without labels, how do we learn representations?
We can learn from context, color, video, sound and invariances.

image