Support vector machines for histogram-based image classification
Support vector machines for histogram-based image classification
复制标题
DOI：
10.1109/72.788646
复制
发表时间：
1999-09-01
影响因子：
--
通讯作者：
Vapnik, VN
中科院分区：
文献类型：
--
作者：
Chapelle, O;Haffner, P;Vapnik, VN
由于特征空间的高维性，传统的分类方法对图像分类任务泛化能力差。本文表明，支持向量机（SVM）可以很好地泛化在高维直方图为特征的困难图像分类问题上。对从Corel库存照片中提取的图像进行分类，评估了形式为K(x,y) = e(-rho)Sigma(i) \x(i)(a) - y(i)(a)\(b)的重尾RBF核，其中a小于或等于1，b小于或等于2，并显示其远远优于传统的多项式或高斯径向基函数（RBF）核。此外，我们观察到输入x(i) -> x(i)(a)的简单重新映射提高了线性支持向量机的性能，使它们在这个问题上成为RBF核的有效替代方案。
Traditional classification approaches generalize poorly on image classification tasks, because of the high dimensionality of the feature space. This paper shows that support vector machines (SVM's) can generalize well on difficult image classification problems where the only features are high dimensional histograms, Heavy-tailed RBF kernels of the form K(x,y) = e(-rho)Sigma(i) \x(i)(a) - y(i)(a)\(b) with a less than or equal to 1 and b less than or equal to 2 are evaluated on the classification of images extracted from the Corel stock photo collection and shown to far outperform traditional polynomial or Gaussian radial basis function (RBF) kernels. Moreover, we observed that a simple remapping of the input x(i) --> x(i)(a) improves the performance of linear SVM's to such an extend that it makes them, for this problem, a valid alternative to RBF kernels.