对抗式域适配迁移学习研究
Research on Adversarial Domain Adaptation

DOI: 10.12677/CSA.2021.1112290, PP. 2853-2861

Keywords: 迁移学习，同构对抗式域自适配，异构对抗式域自适配
Transfer Learning, Homogeneous Adversarial Domain Adaptation, Heterogeneous Adversarial Domain Adaptation

Full-Text   Cite this paper   Add to My Lib

Abstract:

对抗式域自适配以对抗学习的方式最小化源领域中的任务损失和最大化域混肴损失，学习到领域间的共享特征空间，从而进行特征适配(分布式配)，辅助目标领域学习任务，是当前域自适应研究，被广泛应用到在行人重识别、图像分类和情感分析等领域。本文梳理了当前对抗式域适配研究工作，按照源领域和目标领域标签类别空间异同将其分为同构对抗式域自适配和异构对抗式域自适配两种子类型。依次详细介绍边缘分布同构对抗式域自适配、条件分布对抗域自适配、联合分布对抗域自适配和动态分布对抗域自适配、开放集对抗式域自适配、局部对抗式域自适配和通用对抗式域适配六类子域自适配的研究问题、研究思路及主要研究工作。
Adversarial domain adaptation can be used to learn a shared feature space across domains by minimizing the task loss in the source domain and maximizing the domain confusion loss. Then the task in the target domain can get assistance from feature adaptation (distributed adaptation). Now, adversarial domain adaptation is widely used in the fields of pedestrian re-recognition, image classification, sentiment analysis, and so on. This paper summarizes the current research work of adversarial domain adaptation and divides it into two subtypes: homogeneous adversarial domain adaptation, heterogeneous adversarial domain adaptation according to whether the label space of the source domain and the target domain is the same. This paper also introduces in detail the research problems, research ideas, and main research work of six kinds of subdomain adaptation: marginal distribution adversarial domain adaptation, conditional distribution adversarial domain adaptation, joint distribution adversarial domain adaptation, dynamic distributed adversarial domain adaptation, open set adversarial domain adaptation, partial adversarial domain adaptation, and universal adversarial domain adaptation.

References

[1]	Pan, S.J. and Yang, Q. (2010) A Survey on Transfer Learning. IEEE Transactions on Knowledge and Data Engineering, 22, 1345-1359. https://doi.org/10.1109/TKDE.2009.191
[2]	Wang, J., Chen, Y., Feng, W., et al. (2020) Transfer Learning with Dynamic Distribution Adaptation. ACM Transactions on Intelligent Systems and Technology, 11, 1-25. https://doi.org/10.1145/3360309
[3]	Ganin, Y., Ustinova, E., Ajakan, H., et al. (2017) Domain-Adversarial Training of Neural Networks. In: Csurka, G., Ed., Domain Adaptation in Computer Vision Applications, Springer Inter-national Publishing, Cham, 189-209. http://link.springer.com/10.1007/978-3-319-58347-1_10 https://doi.org/10.1007/978-3-319-58347-1_10
[4]	Tzeng, E., Hoffman, J., Saenko, K., et al. (2021) Adversarial Discriminative Domain Adaptation. http://arxiv.org/abs/1702.05464
[5]	戴宏, 盛立杰, 苗启广. 基于胶囊网络的对抗判别域适应算法[J]. 计算机研究与发展, 2021, 58(9): 1997-2012.
[6]	Xu, M., Zhang, J., Ni, B., et al. (2020) Adversarial Domain Adaptation with Domain Mixup: 04. Proceedings of the AAAI Conference on Artificial Intelligence, 34, 6502-6509. https://doi.org/10.1609/aaai.v34i04.6123
[7]	Zhang, Y., Tang, H., Jia, K., et al. (2019) Domain-Symmetric Net-works for Adversarial Domain Adaptation. 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Long Beach, 6-20 June 2019, 5026-5035. https://ieeexplore.ieee.org/document/8953920 https://doi.org/10.1109/CVPR.2019.00517
[8]	Tang, X. and Zhang, X. (2020) Conditional Adversarial Domain Adaptation Neural Network for Motor Imagery EEG Decoding. Entropy, 22, 96. https://doi.org/10.3390/e22010096
[9]	Chen, M., Zhao, S., Liu, H., et al. (2020) Adversarial-Learned Loss for Domain Adaptation: 04. Proceedings of the AAAI Conference on Artificial Intelligence, 34, 3521-3528. https://doi.org/10.1609/aaai.v34i04.5757
[10]	Robbiano, L., Rahman, M.R.U., Galasso, F., et al. (2021) Adversar-ial Branch Architecture Search for Unsupervised Domain Adaptation. http://arxiv.org/abs/2102.06679
[11]	Pei, Z., Cao, Z., Long, M., et al. (2021) Multi-Adversarial Domain Adaptation. Thirty-Second AAAI Conference on Artificial Intelli-gence, New Orleans, 2-7 February 2018, 3934-3941. https://www.aaai.org/ocs/index.php/AAAI/AAAI18/paper/view/17067
[12]	Long, M., Cao, Z., Wang, J., et al. (2018) Conditional Adversarial Domain Adaptation. In: Proceedings of the 32nd International Conference on Neural In-formation Processing Systems, Curran Associates Inc., Red Hook, 1647-1657.
[13]	Li, S., Liu, C.H., Xie, B., et al. (2019) Joint Adversarial Domain Adaptation. Proceedings of the 27th ACM International Conference on Multimedia, Nice, 21-25 October 2019, 729-737. https://doi.org/10.1145/3343031.3351070
[14]	Wang, J., Feng, W., Chen, Y., et al. (2018) Visual Domain Adaptation with Manifold Embedded Distribution Alignment. Proceedings of the 26th ACM International Conference on Multimedia, Seoul, 22-26 October 2018, 402-410. https://doi.org/10.1145/3240508.3240512
[15]	Yu, C., Wang, J., Chen, Y., et al. (2019) Transfer Learning with Dynamic Adversarial Adaptation Network. 2019 IEEE International Conference on Data Mining (ICDM), Beijing, 8-11 November 2019, 778-786. https://doi.org/10.1109/ICDM.2019.00088
[16]	Busto, P.P. and Gall, J. (2017) Open Set Domain Adaptation. 2017 IEEE International Conference on Computer Vision (ICCV), Venice, 22-29 October 2017, 754-763. http://ieeexplore.ieee.org/document/8237350 https://doi.org/10.1109/ICCV.2017.88
[17]	Saito, K., Yamamoto, S., Ushiku, Y., et al. (2018) Open Set Domain Adaptation by Backpropagation. In: Ferrari, V., Hebert, M., Sminchisescu, C., et al., Eds., Computer Vision—ECCV 2018, Springer International Publishing, Cham, 156-171. http://link.springer.com/10.1007/978-3-030-01228-1_10 https://doi.org/10.1007/978-3-030-01228-1_10
[18]	Liu, H., Cao, Z., Long, M., et al. (2019) Separate to Adapt: Open Set Domain Adaptation via Progressive Separation. 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Long Beach, 16-20 June 2019, 2922-2931. https://ieeexplore.ieee.org/document/8953906 https://doi.org/10.1109/CVPR.2019.00304
[19]	Cao, Z., Ma, L., Long, M., et al. (2018) Partial Adversarial Do-main Adaptation. ECCV 2018: 15th European Conference, Munich, 8-14 September 2018, 139-155.
[20]	Cao, Z., Long, M., Wang, J., et al. (2018) Partial Transfer Learning with Selective Adversarial Networks. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City, 18-22 June 2018, 2724-2732. https://ieeexplore.ieee.org/document/8578386 https://doi.org/10.1109/CVPR.2018.00288
[21]	Cao, Z., You, K., Long, M., et al. (2019) Learning to Transfer Examples for Partial Domain Adaptation. 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Long Beach, 16-20 June 2019, 2980-2989. https://ieeexplore.ieee.org/document/8954395 https://doi.org/10.1109/CVPR.2019.00310
[22]	You, K., Long, M., Cao, Z., et al. (2019) Universal Domain Ad-aptation. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Long Beach, 15-20 June 2019, 2720-2729. https://openaccess.thecvf.com/content_CVPR_2019/ html/You_Universal_Domain_Adaptation_CVPR_2019_paper.html
[23]	Fu, B., Cao, Z., Long, M., et al. (2020) Learning to Detect Open Classes for Universal Domain Adaptation. In: Vedaldi, A., Bischof, H., Brox, T., et al., Eds., Computer Vision—ECCV 2020, Springer International Publishing, Cham, 567-583. https://link.springer.com/10.1007/978-3-030-585 55-6_34 https://doi.org/10.1007/978-3-030-58555-6_34

Full-Text