Eprint first made available on ORBi (E-prints, working papers and research blog)
Towards generic image classification: an extensive empirical study
Marée, Raphaël; Geurts, Pierre; Wehenkel, Louis
2014
 

Files


Full Text
maree-Towards-TR2014-full.pdf
Author preprint (3.25 MB)
Download

All documents in ORBi are protected by a user license.

Send to



Details



Abstract :
[en] This paper considers the general problem of image classification without using any prior knowledge about image classes. We study variants of a method based on supervised learning whose common steps are the extraction of random subwindows described by raw pixel intensity values and the use of ensemble of extremely randomized trees to directly classify images or to learn image features. The influence of method parameters and variants is thoroughly evaluated so as to provide baselines and guidelines for future studies. Detailed results are provided on 80 publicly available datasets that depict very diverse types of images (more than 3800 image classes and over 1.5 million images).
Research center :
Giga-Systems Biology and Chemical Biology - ULiège
Disciplines :
Computer science
Author, co-author :
Marée, Raphaël  ;  Université de Liège - ULiège > Electrical Engineering and Computer Science > GIGA-Research
Geurts, Pierre ;  Université de Liège - ULiège > Dép. d'électric., électron. et informat. (Inst.Montefiore) > Algorith. des syst. en interaction avec le monde physique
Wehenkel, Louis  ;  Université de Liège - ULiège > Dép. d'électric., électron. et informat. (Inst.Montefiore) > Systèmes et modélisation
Language :
English
Title :
Towards generic image classification: an extensive empirical study
Publication date :
23 December 2014
Version :
1
Number of pages :
48
Name of the research project :
Cytomine
Funders :
DGRE - Région wallonne. Direction générale des Relations extérieures [BE]
Available on ORBi :
since 24 December 2014

Statistics


Number of views
218 (18 by ULiège)
Number of downloads
341 (13 by ULiège)

Bibliography


Similar publications



Contact ORBi