Multiple classifier combination for face-based identity verification
Multiple classifier combination for face-based identity verification
复制标题
DOI：
10.1016/j.patcog.2004.01.008
复制
发表时间：
2004-07-01
影响因子：
8
通讯作者：
Vandendorpe, L
中科院分区：
文献类型：
--
作者：
Czyz, J;Kittler, J;Vandendorpe, L
当组合来自多个分类器的输出时，有许多组合规则可用。固定组合规则虽然易于实现，但只有在受限条件下才是最优的。在不满足最优性条件的情况下，我们讨论并评价了它们的性能。然后，在基于人脸的身份验证的上下文中，对真实数据上的固定组合规则和可训练组合规则进行比较。通过组合五个不同人脸验证专家的输出来对人脸图像进行分类。结果表明，在XM2VTS数据库上，无论是使用固定的还是可训练的组合规则，错误率都比最好的单一专家降低了50%。(C)2004年模式识别学会。爱思唯尔有限公司出版。保留所有权利。
When combining outputs from multiple classifiers, many combination rules are available. Although easy to implement, fixed combination rules are optimal only in restrictive conditions. We discuss and evaluate their performance when the optimality conditions are not fulfilled. Fixed combination rules are then compared with trainable combination rules on real data in the context of face-based identity verification. The face images are classified by combining the outputs of five different face verification experts. It is demonstrated that a reduction in the error rates of up to 50% over the best single expert is achieved on the XM2VTS database, using either fixed or trainable combination rules. (C) 2004 Pattern Recognition Society. Published by Elsevier Ltd. All rights reserved.