Human Detection via Classification on Riemannian Manifolds

By Oncel Tuzel, Fatih Murat Porikli, Peter Meer. 2007 IEEE Conference on Computer Vision and Pattern Recognition, 2007

550 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/b5702b5726b3dc58d97100061b4f54020631d9e6

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

Boosted local structured HOG-LBP for object localization (2011, methodology): “Impressive performance improvement in object localization has been achieved via the progress in: 1) learning object structure [5, 8, 19, 20, 27] and detector model, and 2) learning low-level feature based appearance model [3, 10, 18 , 21, 24, 25].”
On feature combination for multiclass object classification (2009, methodology): “We use the covariances of simple per-pixel features described in [19] (tangent-space projected).”
Beyond Covariance: Feature Representation with Nonlinear Kernel Matrices (2015, methodology): “A (nonsingular) covariance matrix belongs to the set of symmetric positive-definite (SPD) matrices, which forms a connected Riemannian manifold [23].”
Human detection using partial least squares analysis (2009, methodology): “Previous studies [14, 22, 26] have shown that significant improvement in human detection can be achieved using different types (or combinations) of low-level features.”
Monocular Pedestrian Detection: Survey and Experiments (2009, methodology): “Significant speedups can be obtained by either coupling the sliding window approach with a classifier cascade of increasing complexity [45], [52], [63], [71], [74], [76], [80], [83] or by restricting the search space based on known camera geometry and prior information about the target object class.”
Robust Multiperson Tracking from a Mobile Platform (2009, background): “Human detection has reached an impressive level [10], [16], [30], [53], [54], [57], with many systems also being able to estimate the silhouettes of the detected pedestrians [19], [30], [48], [58].”
Human detection from images and videos: A survey (2016, methodology): “In particular, the comparison was performed with the existing object descriptors using rectangular features in [158], HOG in [22], covariance feature (Cov) in [155], edgelet in [171], generalised Haar + feature Mining (labelled as G.”
Distribution Consistency Based Covariance Metric Networks for Few-Shot Learning (2019, methodology): “Local Covariance Representation The covariance matrix has been widely used as a region descriptor (Tuzel, Porikli, and Meer 2006; 2007; Tabia et al. 2014) or a general representation (Wang et al. 2012; Harandi, Salzmann, and Porikli 2014; Wang et al. 2015b; 2015a; Huang et al. 2015; Wang et al.…”
Recent advances and trends in visual tracking: A review (2011, background): “The feature fusion scheme typically achieves boosted system performance or robustness, which attracts much attention of researchers from multimedia, computer vision and audio–visual speech processing, etc. Tuzel et al. [55] utilized the covariance matrix as the descriptor for human representation.”
Feature representation for statistical-learning-based object detection: A review (2015, methodology): “Covariance features: Covariance features [69] are based on the covariance matrix of eight kinds of local attributes as [ x, y, |Ix|, |Iy|, √ I2 x + I 2 y , |Ixx|, |Iyy|, tan−1(|Iy |/|Ix|) ]T .”
Regionlets for Generic Object Detection (2013, background): “Object detection is arguably an indispensable component for most vision tasks, and it has achieved prominent success for specific targets such as faces [ 1], [18] and pedestrians [3], [19], [20], [21], [22], [23], [24].”
Max-Margin Object Detection (2015, methodology): “Many recent approaches include features for part-based-modeling, methods for combining local features, or dimensionality reduction [6, 20, 16, 4, 18].”
Safety in numbers: Learning categories from few examples with multi model knowledge transfer (2010, methodology): “Specifically, we used four different image descriptors: PHOG Shape Descriptors [2], Appearance Descriptors [14], Region Covariance [23] and Local Binary Patterns.”
Discriminative local binary patterns for human detection in personal album (2008, methodology): “This number for COV is about 2.1, while HOG takes exactly 4.”
Cat Head Detection - How to Effectively Exploit Shape and Texture Features (2008, methodology): “The gradient features are extracted from the edge map [4,3] or oriented gradients, which mainly include SIFT [8], EOH [7], HOG [1], covariance matrix[17], shapelet [15], and edgelet [19].”