Conference item icon

Conference item

What have we learned from deep representations for action recognition?

Abstract:

As the success of deep models has led to their deployment in all areas of computer vision, it is increasingly important to understand how these representations work and what they are capturing. In this paper, we shed light on deep spatiotemporal representations by visualizing what two-stream models have learned in order to recognize actions in video. We show that local detectors for appearance and motion objects arise to form distributed representations for recognizing human actions. Key obse...

Expand abstract
Publication status:
Published
Peer review status:
Peer reviewed
Version:
Accepted manuscript

Actions


Access Document


Files:
Publisher copy:
10.1109/CVPR.2018.00818

Authors


Feichtenhofer, C More by this author
Wildes, RP More by this author
More by this author
Institution:
University of Oxford
Division:
MPLS Division
Department:
Engineering Science
Oxford college:
Brasenose College
Austrian Science Fund More from this funder
Natural Sciences and Engineering Research Council of Canada More from this funder
Canada First Research Excellence Fund More from this funder
Austrian Academy of Sciences More from this funder
Publisher:
Institute of Electrical and Electronics Engineers Publisher's website
Publication date:
2018-12-18
Acceptance date:
2018-02-28
DOI:
Pubs id:
pubs:859251
URN:
uri:4e03a2a0-0124-4cde-bb99-a77ee88664b2
UUID:
uuid:4e03a2a0-0124-4cde-bb99-a77ee88664b2
Local pid:
pubs:859251

Terms of use


Metrics



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

TO TOP