Histogram intersection kernel for image classification

By Annalisa Barla, Francesca Odone, Alessandro Verri. Proceedings 2003 International Conference on Image Processing (Cat. No.03CH37429), 2003

354 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/d6b3fe829050c9bf3e0b1e073e5f4ba3065a0e76

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

In Media Technology

Citations in the corpus

Voxel Cloud Connectivity Segmentation - Supervoxels for Point Clouds (2013, methodology): “Color distance D c , is the euclidean distance in CIELab space, normalized by a constant m. Distance in FPFH space, D f , is calculated using the Histogram Intersection Kernel [2].”
Training highly multiclass classifiers (2014, methodology): “, 1999) on the combined feature representation using the intersection kernel (Barla et al., 2003) to produce a 1024-dimensional or 479-dimensional input vector per image (see Tab.”
Sense discovery via co-clustering on images and text (2015, methodology): “Given this representation, we simply model the likelihood term as the histogram intersection χ(·, ·) [2] of the mean and the input feature vector.”
Bag-of-Visual-Words Scene Classifier With Local and Global Features for High Spatial Resolution Remote Sensing Imagery (2016, methodology): “The HIK has been successfully applied in image classification using color histogram features [15].”
A Color-Texture-Structure Descriptor for High-Resolution Satellite Image Classification (2016, methodology): “To ensure a fair comparison, we use the training set to train an SVM classifier with the HIK and the remaining images for testing; moreover, the parameters are set as recommended by the authors [59], i.”
NormFace: L2 Hypersphere Embedding for Face Verification (2017, methodology): “Moreover, the newly proposed score histogram technique (HIK-SVM in the table) can improve the accuracy further by a signi cant gap.”
Fast Point Feature Histograms (FPFH) for 3D registration (2009, methodology): “To illustrate that the features are discriminative, in the left part of the figure we assembled a confusion matrix with gray values representing the distances between the mean histograms of the different shapes, obtained using the Histogram Intersection Kernel [18]:”
The Fisher Kernel Coding Framework for High Spatial Resolution Scene Classification (2016, methodology): “During the training of the SVM classifier, the histogram intersection kernel (HIK) is adopted due to its performance in image classification [44].”
On Valid Optimal Assignment Kernels and Applications to Graph Classification (2016, background): “This includes structured data like images [2, 5, 11], 3d shapes [1], chemical compounds [8] and proteins [4], which are often represented by graphs.”
Efficient Subwindow Search: A Branch and Bound Framework for Object Localization (2009, neutral): “For 1⁄4 1, we obtain the ordinary histogram intersection measure [20], [21].”
Label Partitioning For Sublinear Ranking (2013, methodology): “Then, KPCA is performed (Schoelkopf et al., 1999) on the combined feature representation using the intersection kernel (Barla et al., 2003) to produce a 1024 dimensional input vector.”
A 2-D wavelet decomposition-based bag-of-visual-words model for land-use scene classification (2014, neutral): “Two well-known kernels – the classical radial basis function (RBF) kernel (Song, Duan, and Jiang 2012) and the histogram intersection (HI) kernel (Barla, Odone, and Verri 2003; Odone, Barla, and Verri 2005) – were taken into account in the study.”
Violence Detection in Video Using Computer Vision Techniques (2011, methodology): “It is wellknown that the choice of SVM kernel can significantly affect performance; in our experiments, we explore the following popular kernels: the histogram intersection kernel [1] (HIK), radial basis function (RBF) and Chi-Squared kernel, which is a variant of RBF that uses χ(2) distance.”
Local Novelty Detection in Multi-class Recognition Problems (2015, methodology): “To measure similarities between histograms, we apply the histogram intersection kernel, a well-known and widely used kernel for histograms obtained from image data [3, 5, 11, 23, 29].”