TY - JOUR
T1 - Inducing semantic segmentation from an example
AU - Schnitman, Yaar
AU - Caspi, Yaron
AU - Cohen-Or, Daniel
AU - Lischinski, Dani
PY - 2006
Y1 - 2006
N2 - Segmenting an image into semantically meaningful parts is a fundamental and challenging task in computer vision. Automatic methods are able to segment an image into coherent regions, but such regions generally do not correspond to complete meaningful parts. In this paper, we show that even a single training example can greatly facilitate the induction of a semantically meaningful segmentation on novel images within the same domain: images depicting the same, or similar, objects in a similar setting. Our approach constructs a non-parametric representation of the example segmentation by selecting patch-based representatives. This allows us to represent complex semantic regions containing a large variety of colors and textures. Given an input image, we first partition it into small homogeneous fragments, and the possible labelings of each fragment are assessed using a robust voting procedure. Graph-cuts optimization is then used to label each fragment in a globally optimal manner.
AB - Segmenting an image into semantically meaningful parts is a fundamental and challenging task in computer vision. Automatic methods are able to segment an image into coherent regions, but such regions generally do not correspond to complete meaningful parts. In this paper, we show that even a single training example can greatly facilitate the induction of a semantically meaningful segmentation on novel images within the same domain: images depicting the same, or similar, objects in a similar setting. Our approach constructs a non-parametric representation of the example segmentation by selecting patch-based representatives. This allows us to represent complex semantic regions containing a large variety of colors and textures. Given an input image, we first partition it into small homogeneous fragments, and the possible labelings of each fragment are assessed using a robust voting procedure. Graph-cuts optimization is then used to label each fragment in a globally optimal manner.
UR - http://www.scopus.com/inward/record.url?scp=33744933693&partnerID=8YFLogxK
U2 - 10.1007/11612704_38
DO - 10.1007/11612704_38
M3 - ???researchoutput.researchoutputtypes.contributiontojournal.conferencearticle???
AN - SCOPUS:33744933693
SN - 0302-9743
VL - 3852 LNCS
SP - 373
EP - 384
JO - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
JF - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
T2 - 7th Asian Conference on Computer Vision, ACCV 2006
Y2 - 13 January 2006 through 16 January 2006
ER -