Sharing features: efficient boosting procedures for multiclass object detection

By Antonio Torralba, Kevin P. Murphy, William T. Freeman. Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2004. CVPR 2004., 2004

734 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/42de22c119f25d303032396b8f7d962f62d6498b

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Building the gist of a scene: the role of global image features in recognition. (2006, background): “Features of openness are shown in black and features of closedness are shown in white. mechanisms underlying human object recognition (Kanwisher, 2003; Grill-Spector and Malach, 2004) and its modeling (Riesenhuber and Poggio, 1999; Ullman et al., 2002; Torralba et al., 2004; Serre et al., 2005).”
Actionness Ranking with Lattice Conditional Ordinal Random Fields (2014, methodology): “A trained AdaBoost classifier [33] is used to measure the degree that the region includes actionness information with local appearance and spatial information.”
A sparse object category model for efficient learning and exhaustive recognition (2005, background): “Many of these are variations on the " Parts and Structure " model introduced by Fischler and Elschlager [8], though the modern instantiations use image fragments and invariant regions [1] [2] [3] [10] [13] [18] [19].”
An Efficient Shared Multi-Class Detection Cascade (2008, background): “But most of this work focuses on detecting instances of one particular object against the background.”
Fully automatic facial feature point detection using Gabor feature based boosted classifiers (2005, background): “GentleBoost seams to converge faster than AdaBoost, and performs better for object detection problems [20].”
Contour-based learning for object detection (2005, methodology): “Each round of boosting searches for a minimum cost J , over all f and a discrete set for θ, and then can fit a and b analytically (see [21]).”
Pedestrian Detection via Classification on Riemannian Manifolds (2008, unknown): “Ç”
A Unified Framework for Multioriented Text Detection and Recognition (2014, methodology): “Moreover, dictionary based correction methods are able to handle partial missing and spurious characters, which often occur for scene text.”
A multi-view probabilistic model for 3D object classes (2009, background): “How do we handle occlusions? How do we deal with intraclass and pose variability of objects? Most of the current researches in object categorization [7, 5, 9, 16, 1, 31] have focused on modeling object intraclass variability in single views (within a small range of planar rotations) or as a mixture ”
Hard negative classes for multiple object detection (2014, background): “This is because the optimization of part filters is time consuming and not suitable for large scale multi-class object detection tasks.”
A survey on face detection in the wild: Past, present and future (2015, unknown): “Onelimitationoftheaboveapproachesisthattheformulationrequiresalinesearchateachweakclassifiertofind theoptimalweights,whichmakesitslowtotrainandhardtodeployfeaturesharing[132].”
A Spectral Regularization Framework for Multi-Task Structure Learning (2007, background): “This problem is important in a variety of applications, ranging from conjoint analysis [12], to object detection in computer vision [18], to multiple microarray data set integration in computational biology [8] – to mention just a few.”
A Dynamic Conditional Random Field Model for Joint Labeling of Object and Scene Classes (2008, methodology): “We employ the joint boosting framework [19] to build a strong classifier H(c, f(xi);ΘΦ) = ∑M m=1 hm(c, f(x t i);ΘΦ).”
Shape Matching and Object Recognition (2006, background): “Recent work on sharing features [Torralba et al., 2004] has 62”