欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
把一类支持向量机应用到人脸相似性学习中，提出了一种快速的人脸相似性学习方法.和标准支持向量机相比较，一类支持向量机的主要特点是只利用相似样本进行训练，减少了数据量，能快速地进行相似性学习.2个实际人脸数据库上的实验结果表明，本方法能够快速地学习到人脸相似性，其运行时间至多是支持向量机算法的三分之一.
英文摘要：
It was introduced one-class support vector machine （ SVM） into similarity learning for face images and then a fast similarity learning method was presented. Compared with the standard SVM, the proposed SVM used only similar samples to train models, which could decrease the number of training samples and im-prove the training speed for similarity learning. Experimental results on two face databases validated that this method could speed up the training procedure for similarity learning. The running time cost of this method was about one-third of that of the relevant SVM.
同期刊论文项目
同项目期刊论文
期刊信息
位置：