Reinforcement of Keypoint Matching by Co-segmentation in Object Retrieval: Face Recognition Case Study
The paper investigates a certain group of problems in visual detection and identification of near-identical the-same-class objects. We focus on difficult problems for which: (1) the intra-class visual differences are comparable to inter-class differences, (2) views of the objects are distorted both photometrically and geometrically, and (3) objects are randomly placed in images of unpredictable contents. Since detection of the-same-person faces in complex images is one of such problems, we use it as the case study for the proposed approach. The approach combines a relatively inexpensive technique of near-duplicate fragment detection with a novel co-segmentation algorithm. Thus, the initial pool of matching candidates can be found quickly (however, with limited precision, i.e. many false positives can be detected). It is shown that the subsequent co-segmentation can effectively reject false positives and accurately extract the matching objects from random backgrounds.
KeywordsObject matching Near-duplicates Co-segmentation Image mapping Face identification
Unable to display preview. Download preview PDF.
- 1.Philbin, J., Chum, O., Isard, M., Sivic, J., Zisserman, A.: Object Retrieval with Large Vocabularies and Fast Spatial Matching. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2007), pp. 1–8 (2007)Google Scholar
- 4.Chum, O., Perdoch, M., Matas, J.: Geometric min-Hashing: Finding a (Thick) Needle in a Haystack. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2009), pp. 17–24 (2009)Google Scholar
- 7.Li, H., Kim, E., Huang, X., He, L.: Object Matching with a Locally Affine-Invariant Constraint. In: IEEE Confernce on Computer Vision and Pattern Recognition (CVPR 2010), pp. 1641–1648 (2010)Google Scholar
- 9.Vicente, S., Rother, C., Kolmogorov, V.: Object Cosegmentation. In: IEEE Confernce on Computer Vision and Pattern Recognition (CVPR 2011), pp. 2217–2224 (2011)Google Scholar
- 10.Yang, D., Śluzek, A.: Co-segmentation by Keypoint Matching: Incorporating Pixel-to-pixel Mapping into MRF. Nanyang Technological University, SCE - unpublished report, Singapore (2010)Google Scholar
- 11.Hochbaum, D.S., Singh, V.: An Efficient Algorithm for Co-segmentation. In: 12th IEEE International Conference on Computer Vision (ICCV 2009), pp. 269–276 (2009)Google Scholar
- 12.Mukherjee, L., Singh, V., Dyer, C.R.: Half-integrality based algorithms for cosegmentation of images. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2009), pp. 2028–2035 (2009)Google Scholar
- 14.Bookstein, F.L.: Principle Warps: Thin Plate Splines and the Decomposition of Deformations. IEEE Trans. PAMI 16, 460–468 (1989)Google Scholar