Face Morphing Attack Detection and Localization Based on Feature-Wise Supervision
Face Morphing Attack Detection and Localization Based on Feature-Wise Supervision
复制标题
基于特征监督的人脸变形攻击检测和定位
DOI：
10.1109/tifs.2022.3212276
复制
发表时间：
2022
影响因子：
6.8
通讯作者：
Min Long
中科院分区：
文献类型：
--
作者：
Le Qin;Fei Peng;Min Long
为了加强人脸识别系统对变形攻击的安全性，人们提出了许多对策。然而，在现有的人脸变形攻击检测(MAD)中，经典的分数级损失训练的深层网络在刻画不同MAS的内在变形模式方面存在不足，也不能直接应用于差异化的MAD场景。为此，本文提出了一种基于特征监督的人脸MA检测与定位方法。该方法基于不同的变形模式构造细粒度的分类损失，并根据不同MAD场景的特点设计基于相似度和基于距离的差别损失。实验结果和分析表明，细粒度分类损失能够定位MAS检测后的局部变形区域，而差分损失能够提高MAD方法对不可见MAS的泛化能力，增强MAD方法对低分辨率、非正面探测人脸图像的鲁棒性。
To strengthen the security of face recognition systems to morphing attacks (MAs), many countermeasures were proposed. However, in the existing face morphing attack detection (MAD), the deep networks trained by classical score-level losses are weak in characterizing the intrinsic morphing patterns of different MAs, and they also cannot be directly applied to differential MAD scenarios. To this end, this paper presents a method for detecting and locating face MAs by the use of feature-wise supervision. It constructs the fine-grained classification loss on the basis of different morphing patterns, and designs the similarity-based and distance-based differential losses according to the properties of differential MAD scenarios. The experimental results and analysis show that the fine-grained classification loss can locate the local morphed areas after detecting MAs, while the differential losses are able to improve the generalization ability of MAD methods to unseen MAs, and can enhance the robustness of MAD methods to low-resolution and non-frontal probe face images.
登录
查看更多内容
DOI：
10.1109/wacv45572.2020.9093488
发表时间：
2020-01
期刊：
2020 IEEE Winter Conference on Applications of Computer Vision (WACV)
影响因子：
--
作者：
S. Venkatesh;Raghavendra Ramachandra;K. Raja;L. Spreeuwers;R. Veldhuis;C. Busch
通讯作者：
S. Venkatesh;Raghavendra Ramachandra;K. Raja;L. Spreeuwers;R. Veldhuis;C. Busch
DOI：
10.1109/btas.2016.7791169
发表时间：
2016-09
期刊：
2016 IEEE 8th International Conference on Biometrics Theory, Applications and Systems (BTAS)
影响因子：
--
作者：
Ramachandra Raghavendra;K. Raja;C. Busch
通讯作者：
Ramachandra Raghavendra;K. Raja;C. Busch
DOI：
10.1109/tifs.2020.3035252
发表时间：
2020-06
影响因子：
6.8
作者：
K. Raja;Matteo Ferrara;Annalisa Franco;L. Spreeuwers;Ilias Batskos;Florens de Wit;M. Gomez-Barrero;U. Scherhag;Daniel Fischer;S. Venkatesh;Jag Mohan Singh;Guoqiang Li;Loïc Bergeron;Sergey Isadskiy;Raghavendra Ramachandra;C. Rathgeb;Dinusha Frings;U. Seidel;F. Knopjes;R. Veldhuis;D. Maltoni;C. Busch
通讯作者：
K. Raja;Matteo Ferrara;Annalisa Franco;L. Spreeuwers;Ilias Batskos;Florens de Wit;M. Gomez-Barrero;U. Scherhag;Daniel Fischer;S. Venkatesh;Jag Mohan Singh;Guoqiang Li;Loïc Bergeron;Sergey Isadskiy;Raghavendra Ramachandra;C. Rathgeb;Dinusha Frings;U. Seidel;F. Knopjes;R. Veldhuis;D. Maltoni;C. Busch
DOI：
10.1109/btas.2018.8698563
发表时间：
2018-10
期刊：
2018 IEEE 9th International Conference on Biometrics Theory, Applications and Systems (BTAS)
影响因子：
--
作者：
N. Damer;Alexandra Mosegui Saladie;Andreas Braun;Arjan Kuijper
通讯作者：
N. Damer;Alexandra Mosegui Saladie;Andreas Braun;Arjan Kuijper
DOI：
10.1109/ijcb52358.2021.9484383
发表时间：
2021-08
期刊：
2021 IEEE International Joint Conference on Biometrics (IJCB)
影响因子：
--
作者：
Zander Blasingame;Chen Liu
通讯作者：
Zander Blasingame;Chen Liu