Template adaptation for face verification and identification
Template adaptation for face verification and identification
复制标题
DOI：
10.1016/j.imavis.2018.09.002
复制
发表时间：
2018-11-01
影响因子：
4.7
通讯作者：
Zisserman, Andrew
中科院分区：
文献类型：
--
作者：
Crosswhite, Nate;Byrne, Jeffrey;Zisserman, Andrew
人脸识别性能评估传统上集中在一对一的验证上，由图像的Labeled Faces in the Wild数据集[1]和视频的YouTubeFaces数据集[2]推广。相比之下，新发布的IJB-A人脸识别数据集[3]将一对多人脸识别的评估与模板上的一对一人脸验证或图像和视频集统一起来。在本文中，我们研究了模板自适应问题，一种形式的迁移学习的一组媒体的模板。对IJB-A的广泛性能评估显示了一个令人惊讶的结果，也许是最简单的模板自适应方法，将深度卷积网络特征与模板特定的线性SVM相结合，远远优于最先进的方法。我们研究了模板大小、负集构造和分类器融合对性能的影响，然后将模板自适应与具有度量学习、2D和3D对齐的卷积网络进行比较。我们意外的结论是，当与模板自适应相结合时，这些其他方法在基于模板的人脸验证和识别的IJB-A上都实现了几乎相同的最高性能。(C)2018 Elsevier B. V.版权所有。
Face recognition performance evaluation has traditionally focused on one-to-one verification, popularized by the Labeled Faces in the Wild data set [1] for imagery and the YouTubeFaces data set [2] for videos. In contrast, the newly released IJB-A face recognition data set [3] unifies evaluation of one-to-many face identification with one-to-one face verification over templates, or sets of imagery and videos for a subject. In this paper, we study the problem of template adaptation, a form of transfer learning to the set of media in a template. Extensive performance evaluations on IJB-A show a surprising result, that perhaps the simplest method of template adaptation, combining deep convolutional network features with template specific linear SVMs, outperforms the state-of-the-art by a wide margin. We study the effects of template size, negative set construction and classifier fusion on performance, then compare template adaptation to convolutional networks with metric learning, 2D and 3D alignment. Our unexpected conclusion is that these other methods, when combined with template adaptation, all achieve nearly the same top performance on IJB-A for template-based face verification and identification. (C) 2018 Elsevier B.V. All rights reserved.