http:\/\/publications.hevs.ch\/index.php\/publications\/show\/1373<\/a><\/p>\nThe paper is written by Adrien Depeursinge, Antonio Foncubierta-Rodriguez, Dimitri Van De Ville and Henning M\u00fcller.<\/p>\n
Abstract\u2014We propose a texture learning approach that exploits<\/strong><\/p>\nlocal organizations of scales and directions. First, linear<\/strong><\/p>\ncombinations of Riesz wavelets are learned using kernel support<\/strong><\/p>\nvector machines. The resulting texture \u201csignatures\u201d are modeling<\/strong><\/p>\noptimal class\u2013wise discriminatory properties. The visualization<\/strong><\/p>\nof the obtained signatures allows verifying the visual relevance<\/strong><\/p>\nof the learned concepts. Second, the local orientations of the<\/strong><\/p>\nsignatures are optimized to maximize their responses, which<\/strong><\/p>\nis carried out analytically and can still be expressed as a<\/strong><\/p>\nlinear combination of the initial steerable Riesz templates. The<\/strong><\/p>\nglobal process is iteratively repeated to obtain final rotation\u2013<\/strong><\/p>\ncovariant texture signatures. Rapid convergence of class\u2013wise<\/strong><\/p>\nsignatures is observed, which demonstrates that the instances<\/strong><\/p>\nare projected into a feature space that leverages the local<\/strong><\/p>\norganizations of scales and directions. Experimental evaluation<\/strong><\/p>\nreveals an average classification accuracies in the range of 97%<\/strong><\/p>\nto 98% for the Outex TC 00010, the Outex TC 00012, and the<\/strong><\/p>\nContrib TC 00000 suite for even orders of the Riesz transform,<\/strong><\/p>\nand suggests high robustness to changes in images orientation<\/strong><\/p>\nand illumination. The proposed framework requires no arbitrary<\/strong><\/p>\nchoices of scales and directions and is expected to perform well<\/strong><\/p>\nin a large range of computer vision applications.<\/strong><\/p>\n\u00a0<\/p>\n
<\/body><\/html><\/p>\n","protected":false},"excerpt":{"rendered":"
The matlab code used in the TIP 2013 paper ‘Rotation-Covariant Texture Learning Using Steerable Riesz Wavelets’ is now available. The code and the paper can be found in\u00a0http:\/\/publications.hevs.ch\/index.php\/publications\/show\/1373 The paper is written by Adrien Depeursinge, Antonio Foncubierta-Rodriguez, Dimitri Van De Ville and Henning M\u00fcller. Abstract\u2014We propose a texture learning approach that exploits local organizations of … <\/p>\n
Continue reading “Matlab code used in texture paper now available”<\/span><\/a><\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"false","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_jetpack_memberships_contains_paid_content":false,"footnotes":""},"categories":[1],"tags":[],"class_list":["post-131","post","type-post","status-publish","format-standard","hentry","category-uncategorized"],"jetpack_featured_media_url":"","jetpack_shortlink":"https:\/\/wp.me\/p8AP2d-27","jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/medgift.hevs.ch\/wordpress\/wp-json\/wp\/v2\/posts\/131","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/medgift.hevs.ch\/wordpress\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/medgift.hevs.ch\/wordpress\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/medgift.hevs.ch\/wordpress\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/medgift.hevs.ch\/wordpress\/wp-json\/wp\/v2\/comments?post=131"}],"version-history":[{"count":0,"href":"https:\/\/medgift.hevs.ch\/wordpress\/wp-json\/wp\/v2\/posts\/131\/revisions"}],"wp:attachment":[{"href":"https:\/\/medgift.hevs.ch\/wordpress\/wp-json\/wp\/v2\/media?parent=131"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/medgift.hevs.ch\/wordpress\/wp-json\/wp\/v2\/categories?post=131"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/medgift.hevs.ch\/wordpress\/wp-json\/wp\/v2\/tags?post=131"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}