Foreground Focus: Unsupervised Learning from Partially Matching Images

Lee, Yong; Grauman, Kristen
November 2009
International Journal of Computer Vision;Nov2009, Vol. 85 Issue 2, p143
Academic Journal
We present a method to automatically discover meaningful features in unlabeled image collections. Each image is decomposed into semi-local features that describe neighborhood appearance and geometry. The goal is to determine for each image which of these parts are most relevant, given the image content in the remainder of the collection. Our method first computes an initial image-level grouping based on feature correspondences, and then iteratively refines cluster assignments based on the evolving intra-cluster pattern of local matches. As a result, the significance attributed to each feature influences an image’s cluster membership, while related images in a cluster affect the estimated significance of their features. We show that this mutual reinforcement of object-level and feature-level similarity improves unsupervised image clustering, and apply the technique to automatically discover categories and foreground regions in images from benchmark datasets.


Related Articles

  • Shape-From-Silhouette Across Time Part I: Theory and Algorithms. Kong-man Cheung; Baker, Simon; Kanade, Takeo // International Journal of Computer Vision;May2005, Vol. 62 Issue 3, p221 

    Shape-From-Silhouette (SFS) is a shape reconstruction method which constructs a 3D shape estimate of an object using silhouette images of the object. The output of a SFS algorithm is known as the Visual Hull (VH). Traditionally SFS is either performed on static objects, or separately at each...

  • Local multiscale blur estimation based on toggle mapping for sharp region extraction. Gillibert, Luc; Chabardès, Théodore; Marcotegui, Beatriz // IET Image Processing;2018, Vol. 12 Issue 12, p2138 

    In this study, a multiscale local blur estimation is proposed based on the existing local focus measure that combines gradient and toggle mapping. This method evaluates the quality of images regardless of their content (not in an autofocus context) and can predict Optical Character Recognition...

  • Benchmarking Image Segmentation Algorithms. Estrada, Francisco; Jepson, Allan // International Journal of Computer Vision;Nov2009, Vol. 85 Issue 2, p167 

    We present a thorough quantitative evaluation of four image segmentation algorithms on images from the Berkeley Segmentation Database. The algorithms are evaluated using an efficient algorithm for computing precision and recall with regard to human ground-truth boundaries. We test each...

  • A Solution of the Dichromatic Model for Multispectral Photometric Invariance. Huynh, Cong Phuoc; Robles-Kelly, Antonio // International Journal of Computer Vision;Oct2010, Vol. 90 Issue 1, p1 

    In this paper, we address the problem of photometric invariance in multispectral imaging making use of an optimisation approach based upon the dichromatic model. In this manner, we cast the problem of recovering the spectra of the illuminant, the surface reflectance and the shading and specular...

  • Generalised Nonlocal Image Smoothing. Pizarro, Luis; Mrázek, Pavel; Didas, Stephan; Grewenig, Sven; Weickert, Joachim // International Journal of Computer Vision;Oct2010, Vol. 90 Issue 1, p62 

    We propose a discrete variational approach for image smoothing consisting of nonlocal data and smoothness constraints that penalise general dissimilarity measures defined on image patches. One of such dissimilarity measures is the weighted L2 distance between patches. In such a case we derive an...

  • Visual Hull with Imprecise Input. Peng He // World Academy of Science, Engineering & Technology;May2010, Issue 41, p733 

    No abstract available.

  • Recovering Occlusion Boundaries from an Image. Hoiem, Derek; Efros, Alexei; Hebert, Martial // International Journal of Computer Vision;Feb2011, Vol. 91 Issue 3, p328 

    Occlusion reasoning is a fundamental problem in computer vision. In this paper, we propose an algorithm to recover the occlusion boundaries and depth ordering of free-standing structures in the scene. Rather than viewing the problem as one of pure image processing, our approach employs cues from...

  • A New Algorithm of Rain (Snow) Removal in Video. Chen Zhen; Shen Jihong // Journal of Multimedia;Apr2013, Vol. 8 Issue 2, p168 

    The images acquired by outdoor vision system in the rain or snow have low contrast and are blurred, and it can cause serious degradation. Traditional rain (snow) removal method is restricted with the intensity, so the effect is not ideal. According to the characteristic of vision system...

  • Interpreting Line Drawings Based on Constraints between Edges and Vertices. Qiu-Ping Cao; Man-Tun Gao; Yi-Kai Shi // International MultiConference of Engineers & Computer Scientists;2007, p1858 

    Extracting 3D information of an object from its 2D projection is a primary objective in contemporary computer vision research and thus has long been studied. In this paper, a new method for reconstructing the 3D framework models of a planar object from its 2D hidden-part-drawn line drawing is...


Read the Article


Sorry, but this item is not currently available from your library.

Try another library?
Sign out of this library

Other Topics