TY - GEN
T1 - Deep Residual Equivariant Mapping for Multi-angle Face Recognition
AU - Liu, Wei
AU - Wu, Lintai
AU - Xu, Yong
AU - Wang, Dan
N1 - Publisher Copyright:
© 2019, Springer Nature Switzerland AG.
PY - 2019
Y1 - 2019
N2 - Face recognition has caught a lot of attention and plenty of valuable methods have been proposed during the past decades. However, because it is hard to learn geometrically invariant representations, existing face recognition methods still perform relatively poorly in conducting multi-angle face recognition. In this paper, we hypothesize that there is an inherent mapping between the frontal and non-frontal faces, and the non-frontal face representations can be converted into the frontal face representations by an equivariant mapping. To carry out the mapping, we propose a Multi-Angle Deep Residual Equivariant Mapping (MADREM) block which adaptively maps the non-frontal face representation to the frontal face representation. It can be considered the MADREM block carry out face alignment and face normalization in the feature space. The residual equivariant mapping block can enhance the discriminative power of the face representations. Finally, we achieve an accuracy of 99.78% on the LFW dataset and 94.25% on CFP-FP dataset based on proposed multiscale-convolution and residual equivariant mapping block.
AB - Face recognition has caught a lot of attention and plenty of valuable methods have been proposed during the past decades. However, because it is hard to learn geometrically invariant representations, existing face recognition methods still perform relatively poorly in conducting multi-angle face recognition. In this paper, we hypothesize that there is an inherent mapping between the frontal and non-frontal faces, and the non-frontal face representations can be converted into the frontal face representations by an equivariant mapping. To carry out the mapping, we propose a Multi-Angle Deep Residual Equivariant Mapping (MADREM) block which adaptively maps the non-frontal face representation to the frontal face representation. It can be considered the MADREM block carry out face alignment and face normalization in the feature space. The residual equivariant mapping block can enhance the discriminative power of the face representations. Finally, we achieve an accuracy of 99.78% on the LFW dataset and 94.25% on CFP-FP dataset based on proposed multiscale-convolution and residual equivariant mapping block.
KW - Face recognition
KW - Feature equivariance
KW - Multiscale convolution
KW - Residual equivariant mapping
UR - https://www.scopus.com/pages/publications/85075550274
U2 - 10.1007/978-3-030-31456-9_16
DO - 10.1007/978-3-030-31456-9_16
M3 - 会议稿件
AN - SCOPUS:85075550274
SN - 9783030314552
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 145
EP - 154
BT - Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings
A2 - Sun, Zhenan
A2 - He, Ran
A2 - Shan, Shiguang
A2 - Feng, Jianjiang
A2 - Guo, Zhenhua
PB - Springer
T2 - 14th Chinese Conference on Biometric Recognition, CCBR 2019
Y2 - 12 October 2019 through 13 October 2019
ER -