Sökning: id:"swepub:oai:DiVA.org:liu-144284" >
ECO: Efficient Conv...
-
Danelljan, Martin,1989-Linköpings universitet,Datorseende,Tekniska fakulteten
(författare)
ECO: Efficient Convolution Operators for Tracking
- Artikel/kapitelEngelska2017
Förlag, utgivningsår, omfång ...
-
Institute of Electrical and Electronics Engineers (IEEE),2017
-
electronicrdacarrier
Nummerbeteckningar
-
LIBRIS-ID:oai:DiVA.org:liu-144284
-
https://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-144284URI
-
https://doi.org/10.1109/CVPR.2017.733DOI
Kompletterande språkuppgifter
-
Språk:engelska
-
Sammanfattning på:engelska
Ingår i deldatabas
Klassifikation
-
Ämneskategori:ref swepub-contenttype
-
Ämneskategori:kon swepub-publicationtype
Anmärkningar
-
Funding Agencies|SSF (SymbiCloud); VR (EMC2) [2016-05543]; SNIC; WASP; Visual Sweden; Nvidia
-
In recent years, Discriminative Correlation Filter (DCF) based methods have significantly advanced the state-of-the-art in tracking. However, in the pursuit of ever increasing tracking performance, their characteristic speed and real-time capability have gradually faded. Further, the increasingly complex models, with massive number of trainable parameters, have introduced the risk of severe over-fitting. In this work, we tackle the key causes behind the problems of computational complexity and over-fitting, with the aim of simultaneously improving both speed and performance. We revisit the core DCF formulation and introduce: (i) a factorized convolution operator, which drastically reduces the number of parameters in the model; (ii) a compact generative model of the training sample distribution, that significantly reduces memory and time complexity, while providing better diversity of samples; (iii) a conservative model update strategy with improved robustness and reduced complexity. We perform comprehensive experiments on four benchmarks: VOT2016, UAV123, OTB-2015, and Temple-Color. When using expensive deep features, our tracker provides a 20-fold speedup and achieves a 13.0% relative gain in Expected Average Overlap compared to the top ranked method [12] in the VOT2016 challenge. Moreover, our fast variant, using hand-crafted features, operates at 60 Hz on a single CPU, while obtaining 65.0% AUC on OTB-2015.
Ämnesord och genrebeteckningar
Biuppslag (personer, institutioner, konferenser, titlar ...)
-
Bhat, GoutamLinköpings universitet,Datorseende,Tekniska fakulteten(Swepub:liu)goubh24
(författare)
-
Khan, Fahad Shahbaz,1983-Linköpings universitet,Datorseende,Tekniska fakulteten(Swepub:liu)fahkh30
(författare)
-
Felsberg, Michael,1974-Linköpings universitet,Datorseende,Tekniska fakulteten(Swepub:liu)micfe03
(författare)
-
Linköpings universitetDatorseende
(creator_code:org_t)
Sammanhörande titlar
-
Ingår i:Proceedings 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR): Institute of Electrical and Electronics Engineers (IEEE), s. 6931-693997815386045719781538604588
Internetlänk
Hitta via bibliotek
Till lärosätets databas