000212937 001__ 212937
000212937 005__ 20190416220325.0
000212937 037__ $$aCONF
000212937 245__ $$aIntroducing Geometry in Active Learning for Image Segmentation
000212937 269__ $$a2015
000212937 260__ $$c2015
000212937 336__ $$aConference Papers
000212937 520__ $$aWe propose an Active Learning approach to training a segmentation classifier that exploits geometric priors to streamline the annotation process in 3D image volumes. To this end, we use these priors not only to select voxels most in need of annotation but to guarantee that they lie on 2D planar patch, which makes it much easier to annotate than if they were randomly distributed in the volume. A simplified version of this approach is effective in natural 2D images. We evaluated our approach on Electron Microscopy and Magnetic Resonance image volumes, as well as on natural images. Comparing our approach against several accepted baselines demonstrates a marked performance increase.
000212937 6531_ $$aActive Learning
000212937 6531_ $$aImage Segmentation
000212937 6531_ $$aBiomedical Imaging
000212937 700__ $$0247782$$aKonyushkova, Ksenia$$g237229
000212937 700__ $$0245861$$aSznitman, Raphael$$g177109
000212937 700__ $$0240252$$aFua, Pascal$$g112366
000212937 7112_ $$ainternational conference in Computer Vision$$cSantiago, Chile$$dDecember 13-16, 2015
000212937 8564_ $$s1718837$$uhttps://infoscience.epfl.ch/record/212937/files/2080.pdf$$yPublisher's version$$zPublisher's version
000212937 8564_ $$s964718$$uhttps://infoscience.epfl.ch/record/212937/files/supplementary.pdf
000212937 909C0 $$0252087$$pCVLAB$$xU10659
000212937 909CO $$ooai:infoscience.tind.io:212937$$pconf$$pIC$$qGLOBAL_SET
000212937 917Z8 $$x237229
000212937 917Z8 $$x237229
000212937 937__ $$aEPFL-CONF-212937
000212937 973__ $$aEPFL$$rREVIEWED$$sACCEPTED
000212937 980__ $$aCONF