Thesis icon

Thesis

Learning dense prediction: from correspondence to segmentation

Abstract:

Dense prediction is the task of predicting a label for each pixel in the image. Given 3D data (point clouds or RGB-D images) as input, dense prediction can also be extended to 3D space and assign each 3D point/location a label. According to the label type, dense prediction can be mainly categorized as depth estimation, motion prediction, segmentation, and other related tasks. There are four major challenges for learning dense predictions: i) how to significantly improve the accuracy and re...

Expand abstract

Actions


Access Document


Files:

Authors


More by this author
Division:
MPLS
Department:
Engineering Science
Role:
Author

Contributors

Institution:
University of Oxford
Division:
MPLS
Department:
Engineering Science
Role:
Supervisor
Institution:
University of Oxford
Division:
MPLS
Department:
Engineering Science
Role:
Supervisor
More from this funder
Name:
European Research Council
Funder identifier:
http://dx.doi.org/10.13039/501100000781
Funding agency for:
Torr, P
Grant:
ERC-2012-AdG 321162-HELIOS
More from this funder
Name:
Engineering and Physical Sciences Research Council
Funding agency for:
Torr, P
Grant:
EP/N019474/1
EP/M013774/1
Programme:
MURI Grant / Seebibyte
More from this funder
Name:
Turing AI Fellowship
Funding agency for:
Torr, P
Grant:
EP/W002981/1
More from this funder
Name:
Snap Inc.
Funding agency for:
Torr, P
More from this funder
Name:
Baidu Inc.
Funding agency for:
Torr, P
Type of award:
DPhil
Level of award:
Doctoral
Awarding institution:
University of Oxford

Terms of use


Views and Downloads






If you are the owner of this record, you can report an update to it here: Report update to this record

TO TOP