English   español  
Por favor, use este identificador para citar o enlazar a este item: http://hdl.handle.net/10261/133257
Compartir / Impacto:
Estadísticas
Add this article to your Mendeley library MendeleyBASE
Visualizar otros formatos: MARC | Dublin Core | RDF | ORE | MODS | METS | DIDL
Título

Discriminative learning of deep convolutional feature point descriptors

AutorSimo-Serra, Edgar; Trulls, Eduard; Ferraz, Luis; Kokkinos, Iasonas; Fua, Pascal; Moreno-Noguer, Francesc
Fecha de publicación2015
EditorInstitute of Electrical and Electronics Engineers
CitaciónICCV 2015
ResumenDeep learning has revolutionalized image-level tasks such as classification, but patch-level tasks, such as correspondence, still rely on hand-crafted features, e.g. SIFT. In this paper we use Convolutional Neural Networks (CNNs) to learn discriminant patch representations and in particular train a Siamese network with pairs of (non-)corresponding patches. We deal with the large number of potential pairs with the combination of a stochastic sampling of the training set and an aggressive mining strategy biased towards patches that are hard to classify. By using the L2 distance during both training and testing we develop 128-D descriptors whose euclidean distances reflect patch similarity, and which can be used as a drop-in replacement for any task involving SIFT. We demonstrate consistent performance gains over the state of the art, and generalize well against scaling and rotation, perspective transformation, non-rigid deformation, and illumination changes. Our descriptors are efficient to compute and amenable to modern GPUs, and are publicly available.
DescripciónTrabajo presentado a la International Conference on Computer Vision celebrada en Santiago (España) del 7 al 13 de diciembre de 2015.
Versión del editorhttp://dx.doi.org/10.1109/ICCV.2015.22
URIhttp://hdl.handle.net/10261/133257
DOI10.1109/ICCV.2015.22
Aparece en las colecciones: (IRII) Comunicaciones congresos
Ficheros en este ítem:
Fichero Descripción Tamaño Formato  
Feature-Point-Descriptors.pdf5,12 MBAdobe PDFVista previa
Visualizar/Abrir
Mostrar el registro completo
 

Artículos relacionados:


NOTA: Los ítems de Digital.CSIC están protegidos por copyright, con todos los derechos reservados, a menos que se indique lo contrario.