Journal article icon

Journal article

Diagnosing and preventing instabilities in recurrent video processing

Abstract:

Recurrent models are a popular choice for video enhancement tasks such as video denoising or super-resolution. In this work, we focus on their stability as dynamical systems and show that they tend to fail catastrophically at inference time on long video sequences. To address this issue, we (1) introduce a diagnostic tool which produces input sequences optimized to trigger instabilities and that can be interpreted as visualizations of temporal receptive fields, and (2) propose two approaches ...

Expand abstract
Publication status:
Published
Peer review status:
Peer reviewed

Actions


Access Document


Files:
Publisher copy:
10.1109/tpami.2022.3160350

Authors


More by this author
Institution:
University of Oxford
Division:
MPLS
Department:
Engineering Science
Role:
Author
Publisher:
IEEE
Journal:
IEEE Transactions on Pattern Analysis and Machine Intelligence More from this journal
Volume:
45
Issue:
2
Pages:
1594-1605
Publication date:
2022-03-17
DOI:
EISSN:
1939-3539
ISSN:
0162-8828
Pmid:
35298375
Language:
English
Keywords:
Pubs id:
1250567
Local pid:
pubs:1250567
Deposit date:
2022-08-01

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