Detecting Pedestrians by Learning Shapelet Features

By Payam Sabzmeydani, Greg Mori. 2007 IEEE Conference on Computer Vision and Pattern Recognition, 2007

470 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/b09d132046f3b21f98206eb514ebfcbd73f32513

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

An HOG-LBP human detector with partial occlusion handling (2009, unknown): “Shapelet: The detector [25] using shapelet features.”
Object Detection in 20 Years: A Survey (2019, background): “vidually on different human parts, e.g. head, legs, and arms. As the increase of computing power, people started to design more complex detection models, and since 2005, gradient-based representation [12, 37, 177, 220, 309] and DPM [15, 37, 54] have become the mainstream of pedestrian detection. In ”
A survey on face detection in the wild: Past, present and future (2015, methodology): “Featuresforface/objectdetection. featuresand RotatedHaar-likefeatures[63] itsvariations Rectangularfeatureswithstructure[59, 64] Haar-likefeaturesonmotionfiltered image[65] Pixel-based Pixelpairs[78] features Controlpointset[79] Binarized Modifiedcensustransform[69] features LBPfeatures[71,72] Local”
Survey of Pedestrian Detection for Advanced Driver Assistance Systems (2010, methodology): “In fact, many of the techniques proposed within the last two years [96], [99], [95], [93], [100], [101], [102] use this benchmark, which makes it feasible to gain insights into the proposed module.”
Multiresolution Models for Object Detection (2010, background): “Popular training algorithms include SVMs [23, 13] and boosting [24, 25].”
Recent advances and trends in visual tracking: A review (2011, background): “Recently, gradient features have been proved advantageous in human detection [14,15].”
Discriminative local binary patterns for human detection in personal album (2008, methodology): “We do not include Shapelet [12] and tree-based approach [16], since these two mainly focus on the learning part and impractical due to their low speed.”
New features and insights for pedestrian detection (2010, background): “All competitive classifiers we know of employ statistical learning techniques to learn the mapping from features to scores (indicating the likelihood of a pedestrian being present)—usually either support vector machines [3, 13, 17, 19, 27] or some variant of boosting [23, 27, 28, 30].”
Informed Haar-Like Features Improve Pedestrian Detection (2014, unknown): “Shapelet[23] gradients AdaBoost INRIA 81.”
Local Transform Features and Hybridization for Accurate Face and Human Detection (2013, methodology): “We compared three proposed human detection methods (the BHOG feature-based human detector (BHOG), the LGP+BHOG feature-based human detector (LGP+BHOG), and the hybrid feature-based human detector (HYBRID)) with other existing human detection methods such as MultiFtr+Motion [44], ChnFtrs [45], FPDW [”
Ten Years of Pedestrian Detection, What Have We Learned? (2014, background): “73% DF X X Haar I Shapelet [10] 91.”
Pedestrian Detection via Classification on Riemannian Manifolds (2008, unknown): “Ç”
Action recognition by learning mid-level motion features (2008, methodology): “We first track and stabilize the video sequences using the method in Sabzmeydani and Mori [22].”
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].”
Co-occurrence Histograms of Oriented Gradients for Pedestrian Detection (2009, methodology): “All the methods use different features and classifiers: Dalal et al. used HOG, and RBF kernel SVM and linear SVM [8]; Gavrila et al. used local receptive fields (LRF) and quadratic SVM [2]; Dollar et al. used Haar wavelet and AdaBoost [20]; Sabzmeydani et al. used shapelet and AdaBoost [15]; and Maj”
Pedestrian detection: A benchmark (2009, methodology): “(3) We benchmark seven algorithms [40, 5, 7, 30, 11, 42, 21], either obtained directly from the original authors or reimplemented in-house.”
Optimized Pedestrian Detection for Multiple and Occluded People (2013, methodology): “Many variants have been proposed, seeking improvements through novel features and/or classifiers [22, 19, 14, 3] or by leveraging feature space transformation methods, e.g. Partial Least Squares (PLS) analysis [20], for feature generation and dimensionality reduction.”
Multi-cue onboard pedestrian detection (2009, background): “Popular classiﬁers are SVMs [26, 5, 13, 20, 19] or boosting [31, 7, 33, 23].”
A Survey of Recent Advances in CNN-based Single Image Crowd Counting and Density Estimation (2017, unknown): “Monolithic detection approaches [25, 51, 94, 28] typically are traditional pedestrian detection methods which train a classifier using features (such as Haar wavelets [95], histogram oriented gradients [25], edgelet [100] and shapelet [80]) extracted from a full body.”