Title:  Fast image classification with reduced multiclass support vector machines
Internal authors: 
Issue Date:  2015
Authors:  Melis, Marco; Piras, Luca; Biggio, Battista; Giacinto, Giorgio; Fumera, Giorgio; Roli, Fabio
International coauthors:  no
Language:  Inglese
Book title:  Image Analysis and Processing – ICIAP 2015 (Part 2)
ISBN:  978-3-319-23233-1
Publisher name:  Springer
Book editors:  Vittorio Murino, Enrico Puppo
Volume:  9280
First page:  78
Last page:  88
Number of pages:  11
Digital Object Identifier (DOI):  http://dx.doi.org/10.1007/978-3-319-23234-8_8
Scopus identifier:  2-s2.0-84944724979
ISI identifier:  WOS:000364991400008
Peer review:  Esperti anonimi
Conference name:  18th International Conference on Image Analysis and Processing, ICIAP 2015
Conference date:  September 7-11, 2015
Conference place:  Genoa, Italy
Abstract:  Image classification is intrinsically a multiclass, nonlinear classification task. Support Vector Machines (SVMs) have been successfully exploited to tackle this problem, using one-vs-one or one-vs-all learning schemes to enable multiclass classification, and kernels designed for image classification to handle nonlinearities. To classify an image at test time, an SVM requires matching it against a small subset of the training data, namely, its support vectors (SVs). In the multiclass case, though, the union of the sets of SVs of each binary SVM may almost correspond to the full training set, potentially yielding an unacceptable computational complexity at test time. To overcome this limitation, in this work we propose a well-principled reduction method that approximates the discriminant function of a multiclass SVM by jointly optimizing the full set of SVs along with their coefficients. We show that our approach is capable of reducing computational complexity up to two orders of magnitude without significantly affecting recognition accuracy, by creating a super-sparse, budgeted set of virtual vectors.
Type: 4.1 Contributo in Atti di convegno

Files in This Item:
File Description Type License  
FastImageClassification_ICIAP2015_printed.pdf  versione editoriale Administrator    Request a copy

Questionnaire and social

Share on: