Master’s dissertation (Dissertations and theses)
Generic image classification: random and convolutional approaches
Begon, Jean-Michel
2014
 

Files


Full Text
MasterThesis_Begon.pdf
Author postprint (56.72 MB)
Download
Annexes
Presentation_Begon.pdf
Publisher postprint (2.62 MB)
Présentation
Download

All documents in ORBi are protected by a user license.

Send to



Details



Keywords :
image classification; machine learning; extremely randomized trees; convolution; random; feature learning
Abstract :
[en] Supervised learning introduces genericity in the field of image classification, thus enabling fast progress in the domain. Genericity does not imply ease-of-use, however, and the best methods in term of accuracy, namely convolutional neural networks, suffer from its lack. In this master thesis, we propose an alternative approach relying on extremely randomized trees and random subwindow extraction combine with elements of the convolutional networks. We explore two modes of utilization of the forest: primarily a direct approach where the forest is the final classifier (ET-DIC) and to a lesser extent, a preprocessing step where the forest is used to build a visual dictionary but where the actual classification is undertaken by a support vector machine (ET-FL). We show that, in both modes, our scheme performs better than without using the convolutional network elements but we are not quite yet reaching their performances. The ET-DIC variant keeps more in the line of classification forest advantages but performs less well as far as accuracy is concerned. This is further highlighted by the remarkable stability of the ET-DIC mode. This stability accounts for the ease-of-use of the method but also prevents elaborated optimization. We were able to score an accuracy of 0.613 whereas the record for this mode without the convolutional network elements was of 0.5367. The ET-FL produces better results at the cost of a greater variability of accuracy due to the loss of the ability to favor the interesting filters and a greater overfitting, consequence of the loss of the ensemble smoothing effect. The accuracies range from 0.55 to 0.7431 depending on the choice of hyper-parameters. The computational cost of both methods is much greater than with a traditional forest, however.
Disciplines :
Computer science
Author, co-author :
Begon, Jean-Michel ;  Université de Liège - ULiège > Form. doct. sc. (info.)
Language :
English
Title :
Generic image classification: random and convolutional approaches
Alternative titles :
[fr] Classification générique d'images : approches aléatoires et convolutionnelles
Defense date :
23 June 2014
Number of pages :
63
Institution :
ULiège - Université de Liège
Degree :
Master en sciences informatiques, à finalité approfondie
Promotor :
Geurts, Pierre ;  Université de Liège - ULiège > Montefiore Institute of Electrical Engineering and Computer Science
Marée, Raphaël  ;  Université de Liège - ULiège > Montefiore Institute of Electrical Engineering and Computer Science
President :
Boigelot, Bernard  ;  Université de Liège - ULiège > Montefiore Institute of Electrical Engineering and Computer Science
Jury member :
Wehenkel, Louis  ;  Université de Liège - ULiège > Montefiore Institute of Electrical Engineering and Computer Science
Van Droogenbroeck, Marc  ;  Université de Liège - ULiège > Montefiore Institute of Electrical Engineering and Computer Science
Available on ORBi :
since 18 January 2016

Statistics


Number of views
92 (14 by ULiège)
Number of downloads
98 (9 by ULiège)

Bibliography


Similar publications



Contact ORBi