Selection of scale-invariant parts for object class recognition

By Gyuri Dorkó, Cordelia Schmid. Proceedings Ninth IEEE International Conference on Computer Vision, 2003

370 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/96d9ab468299fe51a4e14d86d8ea953ccf62b900

How researchers see this paper

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

In Aerospace Engineering

In Computer Vision and Pattern Recognition

Citations in the corpus

A performance evaluation of local descriptors (2005, background): “recognition [21], image retrieval [29, 38], robot localization [40], video data mining [41], building panoramas [4], and recognition of object categories [8, 9, 22, 35].”
Modeling scenes with local descriptors and latent aspects (2005, background): “They were designed to have high degree of invariance, and, as a result, are robust to changes in viewpoint and lighting conditions.”
Generic object recognition with boosting (2006, methodology): “Another object recognition approach was introduced by Dorko and Schmid [6].”
Feature Correspondence Via Graph Matching: Models and Global Optimization (2008, background): “[3,4]) represent images as orderless sets of local appearance descriptors, known as bags of features.”
The Pascal Visual Object Classes Challenge 2006 ( VOC 2006 ) Results (2006, methodology): “The local features we experimented with were Yan Ke's PCA-SIFT [2] and actual patches cut out of images at proper scales in Gaussian pyramids.”
Finding People in Images and Videos (2006, background): “…salient local points or regions for object detection has a long history [Schiele and Crowley 1996a,b,c, Schmid and Mohr 1997, Weber et al. 2000, Lowe 2001, Agarwal and Roth 2002, Fergus et al. 2003, Dorkó and Schmid 2003, Lowe 2004, Opelt et al. 2004, Leibe et al. 2005, Mikolajczyk et al. 2005].”
Pedestrian Detection via Classification on Riemannian Manifolds (2008, neutral): “Ç”
Foreground Focus: Unsupervised Learning from Partially Matching Images (2009, background): “Discriminative feature selection strategies have also been explored to detect features that occur frequently in in-class examples but rarely on the background (Quack et al. 2007; Dorko and Schmid 2003).”
The 2005 PASCAL Visual Object Classes Challenge (2005, background): “In recent years there has been a rapid growth in research, and quite some success, in visual recognition of object classes; examples include [1, 5, 10, 14, 18, 28, 39, 43].”
Creating efficient codebooks for visual recognition (2005, neutral): “For simplicity we formulate categorization as supervised learning over a ‘bag of features’ representation [5, 7, 15].”
A Visual Vocabulary for Flower Classification (2006, methodology): “There has been much recent success in using ‘bag of features’ or ‘bag of visual words’ models for object and scene classification [1, 3, 5, 7, 14, 15, 17].”
Combined Object Categorization and Segmentation With an Implicit Shape Model (2004, background): “Recent work by several researchers has shown considerable promise that this problem may be dealt with by using scale-invariant interest point detectors [9, 17, 8].”
Harvesting Image Databases from the Web (2007, methodology): “The discriminative visual words were obtained based on the likelihood ratio of a visual word occurring in the foreground to background images [7].”
Beyond sliding windows: Object localization by efficient subwindow search (2008, methodology): “Despite the simplicity, variants of this method have proven very successful for object classification in recent years [2, 8, 16, 21, 24].”
Annotator rationales for visual recognition (2011, methodology): “• Mutual Information: uses feature selection to automatically select the k = 100 most discriminative visual words to use as a refined bag-of-words, following the model in [5].”