TY - JOUR
T1 - Unsupervised object discovery
T2 - A comparison
AU - Tuytelaars, Tinne
AU - Lampert, Christoph H.
AU - Blaschko, Matthew B.
AU - Buntine, Wray
PY - 2010/6
Y1 - 2010/6
N2 - The goal of this paper is to evaluate and compare models and methods for learning to recognize basic entities in images in an unsupervised setting. In other words, we want to discover the objects present in the images by analyzing unlabeled data and searching for re-occurring patterns. We experiment with various baseline methods, methods based on latent variable models, as well as spectral clustering methods. The results are presented and compared both on subsets of Caltech256 and MSRC2, data sets that are larger and more challenging and that include more object classes than what has previously been reported in the literature. A rigorous framework for evaluating unsupervised object discovery methods is proposed.
AB - The goal of this paper is to evaluate and compare models and methods for learning to recognize basic entities in images in an unsupervised setting. In other words, we want to discover the objects present in the images by analyzing unlabeled data and searching for re-occurring patterns. We experiment with various baseline methods, methods based on latent variable models, as well as spectral clustering methods. The results are presented and compared both on subsets of Caltech256 and MSRC2, data sets that are larger and more challenging and that include more object classes than what has previously been reported in the literature. A rigorous framework for evaluating unsupervised object discovery methods is proposed.
KW - Evaluation
KW - Object discovery
KW - Unsupervised object recognition
UR - http://www.scopus.com/inward/record.url?scp=77951296653&partnerID=8YFLogxK
U2 - 10.1007/s11263-009-0271-8
DO - 10.1007/s11263-009-0271-8
M3 - Article
SN - 0920-5691
VL - 88
SP - 284
EP - 302
JO - International Journal of Computer Vision
JF - International Journal of Computer Vision
IS - 2
ER -