A Trainable System for Object Detection

By Constantine Papageorgiou, Tomaso A. Poggio. 2000

1,485 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

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

In Computer Vision and Pattern Recognition

Citations in the corpus

What Makes for Effective Detection Proposals? (2015, methodology): “As recently as a few years ago, the most successful approaches to object detection utilised the well known “sliding window” paradigm [1]–[3], in which a computationally ef-ﬁcient classiﬁer tests for object presence in every candidate image window.”
Sharing Visual Features for Multiclass and Multiview Object Detection (2007, background): “Such a classifier can be turned into a detector by sliding it across the image (or image pyramid) and classifying each such local window [26], [16], [1].”
Context-based vision system for place and object recognition (2003, background): “Another difference between our approach to object recognition and previous approaches is that we try to recognize a large set of object types (24) in a natural, unconstrained setting (images are collected by a wearable camera), as opposed to recognizing a small number of classes (such as faces and c”
Efficient Classification for Additive Kernel SVMs (2013, methodology): “This class of kernels, which we refer to as additive kernels, includes widely used kernels for histogram-based image comparison like intersection and chi-squared kernels.”
Open Vocabulary Object Detection with Pseudo Bounding-Box Labels (2021, background): “Object detection [11, 19, 20, 27] is a core task in computer vision that has considerably advanced with the adoption of deep learning and continues to attract significant research effort [26, 31, 33].”
Learning to detect objects in images via a sparse, part-based representation (2004, background): “…approaches to object detection have included raw pixel intensities [7], [8], [9], features obtained via global image transformations [10], [11], and local features such as edge fragments [12], [13], rectangle features [14], Gabor filter-based representations [15], and wavelet features [16].”
Pedestrian detection in crowded scenes (2005, methodology): “Previous approaches to pedestrian detection have used either global models, e.g. using full-body appearance [20] or silhouettes [11] [10] [8]; or an assembly of local feature [23] or part detectors [18] [17].”
Human Detection Using Oriented Histograms of Flow and Appearance (2006, methodology): “A polynomial SVM based pedestrian detector (upright whole-body human detector) using rectified Haar wavelets as input descriptors is described in [17] , with a parts (subwindow) based variant in [16].”
Watching Unlabeled Video Helps Learn New Human Actions from Very Few Labeled Snapshots (2013, background): “, [20]), as well as the use of graphics software to generate images of people (often driven by mocap) with variable clothing and body sizes to train pose estimators [26, 12, 27].”
Finding People in Images and Videos (2006, methodology): “The static (appearance based) detectors shown are: R-HOG – our default R-HOG trained on combined data set; R-HOG (static) – R-HOG trained on the Static Test Set alone, essentially same as in Chapter 4; and Wavelet – our version of the static Haar wavelet based detector of Papageorgiou and Poggio [20”
Utility data annotation with Amazon Mechanical Turk (2008, background): “It is obvious from this figure that hips are much more difficult to localize compared to shoulders, knees, elbows, wrists, ankles, the head and the neck.”
Contextual Priming for Object Detection (2003, background): “P ( j on; ~vC) = PM i=1 biG( ; i;Si)G(~vC ;~vi;Vi) PM i=1 biG(~vC ;~vi;Vi) (28)”
An Experimental Study on Pedestrian Classification (2006, methodology): “Papageorgiou and Poggio [5] pioneered the use of overcomplete sets of (Haar) wavelet features in combination with a Support Vector Machine (SVM).”
Support vector tracking (2004, neutral): “Finally, several researchers [22], [9], [13], [14], [15], [20] perform object detection on every frame, ignoring object tracking all together.”
Human Detection via Classification on Riemannian Manifolds (2007, methodology): “In [16], a polynomial support vector machine (SVM) was learned using Haar wavelets as human descriptors.”
Cat Head Detection - How to Effectively Exploit Shape and Texture Features (2008, methodology): “The image features are directly extracted from the image, such as intensity values [14], image patch [6], PCA coefficients [11], and wavelet coefficients [12,16,18].”
Co-occurrence Histograms of Oriented Gradients for Pedestrian Detection (2009, methodology): “used SVM-based parts detectors with Haar wavelet feature and integrated them with SVM [5], [6].”
Monocular Pedestrian Detection: Survey and Experiments (2009, background): “The computational costs are often too high to allow for real-time processing [11], [12], [48], [53], [60], [68].”