Show simple item record

dc.contributor.authorPérez-Pellitero, Eduardo
dc.contributor.authorSalvador, Jordi
dc.contributor.authorRuiz Hidalgo, Javier
dc.contributor.authorRosenhahn, Bodo
dc.contributor.otherUniversitat Politècnica de Catalunya. Departament de Teoria del Senyal i Comunicacions
dc.identifier.citationPérez-Pellitero, E., Salvador, J., Ruiz-Hidalgo, J., Rosenhahn, B. Half hypersphere confinement for piecewise linear regression. A: IEEE Winter Conference on Applications of Computer Vision. "2016 IEEE Winter Conference on Applications of Computer Vision: WACV 2016: Lake Placid, New York, USA: 7-10 March 2016". Lake Placid, NY: Institute of Electrical and Electronics Engineers (IEEE), 2016, p. 1-9.
dc.description.abstractRecent research in piecewise linear regression for Super-Resolution has shown the positive impact of training regressors with densely populated clusters whose datapoints are tight in the Euclidean space. In this paper we further research how to improve the locality condition during the training of regressors and how to better select them during testing time. We study the characteristics of the metrics best suited for the piecewise regression algorithms, in which comparisons are usually made between normalized vectors that lie on the unitary hypersphere. Even though Euclidean distance has been widely used for this purpose, it is suboptimal since it does not handle antipodal points (i.e. diametrically opposite points) properly, as vectors with same module and angle but opposite directions are, for linear regression purposes, identical. Therefore, we propose the usage of antipodally invariant metrics and introduce the Half Hypersphere Confinement (HHC), a fast alternative to Multidimensional Scaling (MDS) that allows to map antipodally invariant distances in the Euclidean space with very little approximation error By doing so, we enable the usage of fast search structures based on Euclidean distances without undermining their speed gains with complex distance transformations. The performance of our method, which we named HHC Regression (HHCR), applied to SuperResolution (SR) improves both in quality (PSNR) and it is faster than any other state-of-the-art method. Additionally, under an application-agnostic interpretation of our regression framework, we also test our algorithm for denoising and depth upscaling with promising results.
dc.format.extent9 p.
dc.publisherInstitute of Electrical and Electronics Engineers (IEEE)
dc.subjectÀrees temàtiques de la UPC::Enginyeria de la telecomunicació::Processament del senyal::Processament de la imatge i del senyal vídeo
dc.subject.lcshImage processing
dc.subject.lcshComputer vision
dc.subject.otherImage denoising
dc.subject.otherImage resolution
dc.subject.otherPiecewise linear techniques
dc.subject.otherRegression analysis – vectors
dc.subject.otherHalf hypersphere confinement
dc.subject.otherPiecewise linear regression
dc.subject.otherDensely populated clusters
dc.subject.otherEuclidean space
dc.subject.otherNormalized vectors
dc.subject.otherUnitary hypersphere
dc.subject.otherAntipodal points
dc.subject.otherDiametrically opposite points
dc.subject.otherAntipodally invariant metrics
dc.subject.otherAntipodally invariant distances
dc.subject.otherHHC regression
dc.subject.otherDepth upscaling
dc.titleHalf hypersphere confinement for piecewise linear regression
dc.typeConference report
dc.subject.lemacImatges -- Processament
dc.subject.lemacVisió per ordinador
dc.contributor.groupUniversitat Politècnica de Catalunya. GPI - Grup de Processament d'Imatge i Vídeo
dc.description.peerreviewedPeer Reviewed
dc.rights.accessRestricted access - publisher's policy
dc.description.versionPostprint (author's final draft)
upcommons.citation.authorPérez-Pellitero, E., Salvador, J., Ruiz-Hidalgo, J., Rosenhahn, B.
upcommons.citation.contributorIEEE Winter Conference on Applications of Computer Vision
upcommons.citation.pubplaceLake Placid, NY
upcommons.citation.publicationName2016 IEEE Winter Conference on Applications of Computer Vision: WACV 2016: Lake Placid, New York, USA: 7-10 March 2016

Files in this item


This item appears in the following Collection(s)

Show simple item record

All rights reserved. This work is protected by the corresponding intellectual and industrial property rights. Without prejudice to any existing legal exemptions, reproduction, distribution, public communication or transformation of this work are prohibited without permission of the copyright holder