An Exemplar-Based CRF for Multi-instance Object Segmentation
Authors:
- Xuming HeXHXuming He
- Stephen GouldSGStephen Gould
- 1 June 2014
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 10636919,p. 296-303
Abstract
We address the problem of joint detection and segmentation of multiple object instances in an image, a key step towards scene understanding. Inspired by data-driven methods, we propose an exemplar-based approach to the task of instance segmentation, in which a set of reference image/shape masks is used to find multiple objects. We design a novel CRF framework that jointly models object appearance, shape deformation, and object occlusion. To tackle the challenging MAP inference problem, we derive an alternating procedure that interleaves object segmentation and shape/appearance adaptation. We evaluate our method on two datasets with instance labels and show promising results.Keywords
This publication has 21 references indexed in Scilit:
- Human Pose Estimation Using a Joint Pixel-wise and Part-wise Formulation2013
- Finding Things: Image Parsing with Regions and Per-Exemplar Detectors2013
- Figure-ground segmentation by transferring window masks2012
- Hough Regions for Joining Instance Localization and Segmentation2012
- Extracting foreground masks towards object recognition2011
- Ensemble of exemplar-SVMs for object detection and beyond2011
- LabelMe: A Database and Web-Based Tool for Image AnnotationInternational Journal of Computer Vision, 2007
- Optimizing Binary MRFs via Extended Roof Duality2007
- Contour-based learning for object detection2005
- "GrabCut"2004