Learning discriminative local binary patterns for face recognition
Learning discriminative local binary patterns for face recognition
复制标题
DOI：
10.1109/fg.2011.5771444
复制
发表时间：
2011-03
期刊：
影响因子：
--
通讯作者：
Daniel Maturana;D. Mery;Á. Soto
中科院分区：
文献类型：
--
作者：
Daniel Maturana;D. Mery;Á. Soto
局部二值模式直方图（LBP）及其变体是用于人脸识别的流行的局部视觉描述符。到目前为止，LBP的大多数变体都是手工设计的，或者是用非监督方法学习的。在这项工作中，我们提出了一个简单的方法来学习判别LBP在监督的方式。该方法将LBP类描述符表示为邻域内的一组像素比较，并以最大化所得直方图的Fisher可分性准则的方式搜索一组像素比较。在标准人脸识别数据集上的测试表明，该方法可以创建紧凑而有区别的描述符。
Histograms of Local Binary Patterns (LBPs) and variations thereof are a popular local visual descriptor for face recognition. So far, most variations of LBP are designed by hand or are learned with non-supervised methods. In this work we propose a simple method to learn discriminative LBPs in a supervised manner. The method represents an LBP-like descriptor as a set of pixel comparisons within a neighborhood and heuristically seeks for a set of pixel comparisons so as to maximize a Fisher separability criterion for the resulting histograms. Tests on standard face recognition datasets show that this method can create compact yet discriminative descriptors.