LARNet: Lie Algebra Residual Network for Face Recognition

Xiaolong Yang, Xiaohong Jia, Dihong Gong, Dong-Ming Yan, Zhifeng Li, Wei Liu
Proceedings of the 38th International Conference on Machine Learning, PMLR 139:11738-11750, 2021.

Abstract

Face recognition is an important yet challenging problem in computer vision. A major challenge in practical face recognition applications lies in significant variations between profile and frontal faces. Traditional techniques address this challenge either by synthesizing frontal faces or by pose invariant learning. In this paper, we propose a novel method with Lie algebra theory to explore how face rotation in the 3D space affects the deep feature generation process of convolutional neural networks (CNNs). We prove that face rotation in the image space is equivalent to an additive residual component in the feature space of CNNs, which is determined solely by the rotation. Based on this theoretical finding, we further design a Lie Algebraic Residual Network (LARNet) for tackling pose robust face recognition. Our LARNet consists of a residual subnet for decoding rotation information from input face images, and a gating subnet to learn rotation magnitude for controlling the strength of the residual component contributing to the feature learning process. Comprehensive experimental evaluations on both frontal-profile face datasets and general face recognition datasets convincingly demonstrate that our method consistently outperforms the state-of-the-art ones.

Cite this Paper

BibTeX
@InProceedings{pmlr-v139-yang21d, title = {LARNet: Lie Algebra Residual Network for Face Recognition}, author = {Yang, Xiaolong and Jia, Xiaohong and Gong, Dihong and Yan, Dong-Ming and Li, Zhifeng and Liu, Wei}, booktitle = {Proceedings of the 38th International Conference on Machine Learning}, pages = {11738--11750}, year = {2021}, editor = {Meila, Marina and Zhang, Tong}, volume = {139}, series = {Proceedings of Machine Learning Research}, month = {18--24 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v139/yang21d/yang21d.pdf}, url = {https://proceedings.mlr.press/v139/yang21d.html}, abstract = {Face recognition is an important yet challenging problem in computer vision. A major challenge in practical face recognition applications lies in significant variations between profile and frontal faces. Traditional techniques address this challenge either by synthesizing frontal faces or by pose invariant learning. In this paper, we propose a novel method with Lie algebra theory to explore how face rotation in the 3D space affects the deep feature generation process of convolutional neural networks (CNNs). We prove that face rotation in the image space is equivalent to an additive residual component in the feature space of CNNs, which is determined solely by the rotation. Based on this theoretical finding, we further design a Lie Algebraic Residual Network (LARNet) for tackling pose robust face recognition. Our LARNet consists of a residual subnet for decoding rotation information from input face images, and a gating subnet to learn rotation magnitude for controlling the strength of the residual component contributing to the feature learning process. Comprehensive experimental evaluations on both frontal-profile face datasets and general face recognition datasets convincingly demonstrate that our method consistently outperforms the state-of-the-art ones.} }
Endnote
%0 Conference Paper %T LARNet: Lie Algebra Residual Network for Face Recognition %A Xiaolong Yang %A Xiaohong Jia %A Dihong Gong %A Dong-Ming Yan %A Zhifeng Li %A Wei Liu %B Proceedings of the 38th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2021 %E Marina Meila %E Tong Zhang %F pmlr-v139-yang21d %I PMLR %P 11738--11750 %U https://proceedings.mlr.press/v139/yang21d.html %V 139 %X Face recognition is an important yet challenging problem in computer vision. A major challenge in practical face recognition applications lies in significant variations between profile and frontal faces. Traditional techniques address this challenge either by synthesizing frontal faces or by pose invariant learning. In this paper, we propose a novel method with Lie algebra theory to explore how face rotation in the 3D space affects the deep feature generation process of convolutional neural networks (CNNs). We prove that face rotation in the image space is equivalent to an additive residual component in the feature space of CNNs, which is determined solely by the rotation. Based on this theoretical finding, we further design a Lie Algebraic Residual Network (LARNet) for tackling pose robust face recognition. Our LARNet consists of a residual subnet for decoding rotation information from input face images, and a gating subnet to learn rotation magnitude for controlling the strength of the residual component contributing to the feature learning process. Comprehensive experimental evaluations on both frontal-profile face datasets and general face recognition datasets convincingly demonstrate that our method consistently outperforms the state-of-the-art ones.
APA
Yang, X., Jia, X., Gong, D., Yan, D., Li, Z. & Liu, W.. (2021). LARNet: Lie Algebra Residual Network for Face Recognition. Proceedings of the 38th International Conference on Machine Learning, in Proceedings of Machine Learning Research 139:11738-11750 Available from https://proceedings.mlr.press/v139/yang21d.html.

Related Material