Multi-Adversarial Domain Adaptation

AAAI, 2018.

Cited by: 125|Bibtex|Views23|Links
EI
Keywords:
deep networktransferable featureDeep Domain ConfusionReverse Gradientdatum distributionMore(20+)
Weibo:
The encouraging results highlight the importance of multi-adversarial domain adaptation in deep neural networks, and suggest that multi-adversarial domain adaptation is able to learn more transferable representations for effective domain adaptation

Abstract:

Recent advances in deep domain adaptation reveal that adversarial learning can be embedded into deep networks to learn transferable features that reduce distribution discrepancy between the source and target domains. Existing domain adversarial adaptation methods based on single domain discriminator only align the source and target data...More

Code:

Data:

0
Introduction
  • Deep networks, when trained on large-scale labeled datasets, can learn transferable representations which are generically useful across diverse tasks and application domains (Donahue et al 2014; Yosinski et al 2014).
  • There is strong motivation to establishing effective algorithms to reduce the labeling consumption by leveraging readily-available labeled data from a different but related source domain.
  • This promising transfer learning paradigm, suffers from the shift in data distributions across different domains, which poses a major obstacle in adapting classification models to target tasks (Pan and Yang 2010).
  • The latest advances have been achieved by embedding domain adaptation modules in the pipeline of deep feature learning to extract domain-invariant representations (Tzeng et al 2014; Long et al 2015; Ganin and Lempitsky 2015; Tzeng et al 2015; Long et al 2016; Bousmalis et al 2016; Long et al 2017)
Highlights
  • Deep networks, when trained on large-scale labeled datasets, can learn transferable representations which are generically useful across diverse tasks and application domains (Donahue et al 2014; Yosinski et al 2014)
  • Recent studies have revealed that deep neural networks can learn more transferable features for domain adaptation (Donahue et al 2014; Yosinski et al 2014), by disentangling explanatory factors of variations behind domains
  • We present a multi-adversarial domain adaptation (MADA) approach, which captures multimode structures to enable fine-grained alignment of different data distributions based on multiple domain discriminators
  • The classification accuracy results on the Office-31 dataset for unsupervised domain adaptation based on AlexNet and ResNet are shown in Table 1
  • The encouraging results highlight the importance of multi-adversarial domain adaptation in deep neural networks, and suggest that multi-adversarial domain adaptation is able to learn more transferable representations for effective domain adaptation
  • This paper presented a novel multi-adversarial domain adaptation approach to enable effective deep transfer learning
Methods
  • The authors compare the proposed multi-adversarial domain adaptation (MADA) with both shallow and deep transfer learning methods: Transfer Component Analysis (TCA) (Pan et al 2011), Geodesic Flow Kernel (GFK) (Gong et al 2012), Deep Domain Confusion (DDC) (Tzeng et al 2014), Deep Adaptation Network (DAN) (Long et al 2015), Residual Transfer Network (RTN) (Long et al 2016), and Reverse Gradient (RevGrad) (Ganin and Lempitsky 2015).
  • RevGrad enables domain adversarial learning (Goodfellow et al 2014) by adapting a single layer of deep networks, which matches the source and target domains by making them indistinguishable for a domain discriminator
Results
  • The classification accuracy results on the Office-31 dataset for unsupervised domain adaptation based on AlexNet and ResNet are shown in Table 1.
  • C → the author Cs → P P → C Avg. ResNet (He et al 2016) DAN (Long et al 2015) RTN (Long et al 2016) RevGrad (Ganin and Lempitsky 2015).
  • MADA outperforms all comparison methods on most transfer tasks.
  • As reported in Table 2, the MADA approach outperforms the comparison methods on most transfer tasks.
  • The encouraging results highlight the importance of multi-adversarial domain adaptation in deep neural networks, and suggest that MADA is able to learn more transferable representations for effective domain adaptation
Conclusion
  • This paper presented a novel multi-adversarial domain adaptation approach to enable effective deep transfer learning.
  • Unlike previous domain adversarial adaptation methods that only match the feature distributions across domains without exploiting the complex multimode structures, the proposed approach further exploits the discriminative structures to enable fine-grained distribution alignment in a multi-adversarial adaptation framework, which can simultaneously promote positive transfer and circumvent negative transfer.
  • Experiments show state of the art results of the proposed approach
Summary
  • Introduction:

    Deep networks, when trained on large-scale labeled datasets, can learn transferable representations which are generically useful across diverse tasks and application domains (Donahue et al 2014; Yosinski et al 2014).
  • There is strong motivation to establishing effective algorithms to reduce the labeling consumption by leveraging readily-available labeled data from a different but related source domain.
  • This promising transfer learning paradigm, suffers from the shift in data distributions across different domains, which poses a major obstacle in adapting classification models to target tasks (Pan and Yang 2010).
  • The latest advances have been achieved by embedding domain adaptation modules in the pipeline of deep feature learning to extract domain-invariant representations (Tzeng et al 2014; Long et al 2015; Ganin and Lempitsky 2015; Tzeng et al 2015; Long et al 2016; Bousmalis et al 2016; Long et al 2017)
  • Objectives:

    The goal of this paper is to design a deep neural network that enables learning of transfer features f = Gf (x) and adaptive classifier y = Gy (f ) to reduce the shifts in the joint distributions across domains, such that the target risk Pr(x,y)∼q [Gy (Gf (x)) = y] minimized by jointly minimizing source risk and distribution discrepancy by multiadversarial domain adaptation
  • Methods:

    The authors compare the proposed multi-adversarial domain adaptation (MADA) with both shallow and deep transfer learning methods: Transfer Component Analysis (TCA) (Pan et al 2011), Geodesic Flow Kernel (GFK) (Gong et al 2012), Deep Domain Confusion (DDC) (Tzeng et al 2014), Deep Adaptation Network (DAN) (Long et al 2015), Residual Transfer Network (RTN) (Long et al 2016), and Reverse Gradient (RevGrad) (Ganin and Lempitsky 2015).
  • RevGrad enables domain adversarial learning (Goodfellow et al 2014) by adapting a single layer of deep networks, which matches the source and target domains by making them indistinguishable for a domain discriminator
  • Results:

    The classification accuracy results on the Office-31 dataset for unsupervised domain adaptation based on AlexNet and ResNet are shown in Table 1.
  • C → the author Cs → P P → C Avg. ResNet (He et al 2016) DAN (Long et al 2015) RTN (Long et al 2016) RevGrad (Ganin and Lempitsky 2015).
  • MADA outperforms all comparison methods on most transfer tasks.
  • As reported in Table 2, the MADA approach outperforms the comparison methods on most transfer tasks.
  • The encouraging results highlight the importance of multi-adversarial domain adaptation in deep neural networks, and suggest that MADA is able to learn more transferable representations for effective domain adaptation
  • Conclusion:

    This paper presented a novel multi-adversarial domain adaptation approach to enable effective deep transfer learning.
  • Unlike previous domain adversarial adaptation methods that only match the feature distributions across domains without exploiting the complex multimode structures, the proposed approach further exploits the discriminative structures to enable fine-grained distribution alignment in a multi-adversarial adaptation framework, which can simultaneously promote positive transfer and circumvent negative transfer.
  • Experiments show state of the art results of the proposed approach
Tables
  • Table1: Accuracy (%) on Office-31 for unsupervised domain adaptation (AlexNet and ResNet)
  • Table2: Accuracy (%) on ImageCLEF-DA for unsupervised domain adaptation (AlexNet and ResNet)
  • Table3: Accuracy (%) on Office-31 for domain adaptation from 31 classes to 25 classes (AlexNet)
Download tables as Excel
Related work
Funding
  • This work was supported by the National Key Research and Development Program of China (2016YFB1000701), National Natural Science Foundation of China (61772299, 61325008, 61502265, 61672313) and Tsinghua National Laboratory (TNList) Key Project
Reference
  • Arjovsky, M., and Bottou, L. 2017. Towards principled methods for training generative adversarial networks. In ICLR.
    Google ScholarFindings
  • Arjovsky, M.; Chintala, S.; and Bottou, L. 2017. Wasserstein gan. arXiv preprint arXiv:1701.07875.
    Findings
  • Ben-David, S.; Blitzer, J.; Crammer, K.; Kulesza, A.; Pereira, F.; and Vaughan, J. W. 2010. A theory of learning from different domains. Machine Learning 79(1-2):151–175.
    Google ScholarLocate open access versionFindings
  • Bengio, Y.; Courville, A.; and Vincent, P. 2013. Representation learning: A review and new perspectives. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) 35(8):1798– 1828.
    Google ScholarLocate open access versionFindings
  • Bousmalis, K.; Trigeorgis, G.; Silberman, N.; Krishnan, D.; and Erhan, D. 2016. Domain separation networks. In NIPS, 343–351.
    Google ScholarLocate open access versionFindings
  • Che, T.; Li, Y.; Jacob, A. P.; Bengio, Y.; and Li, W. 2017. Mode regularized generative adversarial networks. ICLR.
    Google ScholarLocate open access versionFindings
  • Collobert, R.; Weston, J.; Bottou, L.; Karlen, M.; Kavukcuoglu, K.; and Kuksa, P. 2011. Natural language processing (almost) from scratch. Journal of Machine Learning Research (JMLR) 12:2493– 2537.
    Google ScholarLocate open access versionFindings
  • Donahue, J.; Jia, Y.; Vinyals, O.; Hoffman, J.; Zhang, N.; Tzeng, E.; and Darrell, T. 2014. Decaf: A deep convolutional activation feature for generic visual recognition. In ICML.
    Google ScholarFindings
  • Duan, L.; Tsang, I. W.; and Xu, D. 2012. Domain transfer multiple kernel learning. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) 34(3):465–479.
    Google ScholarLocate open access versionFindings
  • Durugkar, I.; Gemp, I.; and Mahadevan, S. 2017. Generative multiadversarial networks. ICLR.
    Google ScholarLocate open access versionFindings
  • Ganin, Y., and Lempitsky, V. 2015. Unsupervised domain adaptation by backpropagation. In ICML.
    Google ScholarFindings
  • Glorot, X.; Bordes, A.; and Bengio, Y. 2011. Domain adaptation for large-scale sentiment classification: A deep learning approach. In ICML.
    Google ScholarFindings
  • Gong, B.; Shi, Y.; Sha, F.; and Grauman, K. 2012. Geodesic flow kernel for unsupervised domain adaptation. In CVPR.
    Google ScholarFindings
  • Goodfellow, I.; Pouget-Abadie, J.; Mirza, M.; Xu, B.; Warde-Farley, D.; Ozair, S.; Courville, A.; and Bengio, Y. 20Generative adversarial nets. In NIPS.
    Google ScholarFindings
  • Gretton, A.; Borgwardt, K.; Rasch, M.; Scholkopf, B.; and Smola, A. 2012. A kernel two-sample test. Journal of Machine Learning Research (JMLR) 13:723–773.
    Google ScholarLocate open access versionFindings
  • He, K.; Zhang, X.; Ren, S.; and Sun, J. 20Deep residual learning for image recognition. In CVPR.
    Google ScholarFindings
  • Hoffman, J.; Guadarrama, S.; Tzeng, E.; Hu, R.; Donahue, J.; Girshick, R.; Darrell, T.; and Saenko, K. 2014. LSDA: Large scale detection through adaptation. In NIPS.
    Google ScholarFindings
  • Jia, Y.; Shelhamer, E.; Donahue, J.; Karayev, S.; Long, J.; Girshick, R.; Guadarrama, S.; and Darrell, T. 2014. Caffe: Convolutional architecture for fast feature embedding. In ACM MM.
    Google ScholarLocate open access versionFindings
  • Krizhevsky, A.; Sutskever, I.; and Hinton, G. E. 2012. Imagenet classification with deep convolutional neural networks. In NIPS.
    Google ScholarFindings
  • Long, M.; Cao, Y.; Wang, J.; and Jordan, M. I. 2015. Learning transferable features with deep adaptation networks. In ICML.
    Google ScholarFindings
  • Long, M.; Zhu, H.; Wang, J.; and Jordan, M. I. 2016. Unsupervised domain adaptation with residual transfer networks. In NIPS, 136– 144.
    Google ScholarLocate open access versionFindings
  • Long, M.; Zhu, H.; Wang, J.; and Jordan, M. I. 2017. Deep transfer learning with joint adaptation networks. In ICML.
    Google ScholarFindings
  • Mansour, Y.; Mohri, M.; and Rostamizadeh, A. 2009. Domain adaptation: Learning bounds and algorithms. In COLT.
    Google ScholarFindings
  • Metz, L.; Poole, B.; Pfau, D.; and Sohl-Dickstein, J. 2017. Unrolled generative adversarial networks. ICLR.
    Google ScholarLocate open access versionFindings
  • Mirza, M., and Osindero, S. 2014. Conditional generative adversarial nets. arXiv preprint arXiv:1411.1784.
    Findings
  • Ngiam, J.; Khosla, A.; Kim, M.; Nam, J.; Lee, H.; and Ng, A. Y. 2011. Multimodal deep learning. In ICML.
    Google ScholarFindings
  • Oquab, M.; Bottou, L.; Laptev, I.; and Sivic, J. 2013. Learning and transferring mid-level image representations using convolutional neural networks. In CVPR.
    Google ScholarFindings
  • Pan, S. J., and Yang, Q. 2010. A survey on transfer learning. IEEE Transactions on Knowledge and Data Engineering (TKDE) 22(10):1345–1359.
    Google ScholarLocate open access versionFindings
  • Pan, S. J.; Tsang, I. W.; Kwok, J. T.; and Yang, Q. 2011. Domain adaptation via transfer component analysis. IEEE Transactions on Neural Networks (TNN) 22(2):199–210.
    Google ScholarLocate open access versionFindings
  • Russakovsky, O.; Deng, J.; Su, H.; Krause, J.; Satheesh, S.; Ma, S.; Huang, Z.; Karpathy, A.; Khosla, A.; Bernstein, M.; Berg, A. C.; and Fei-Fei, L. 2014. ImageNet Large Scale Visual Recognition Challenge.
    Google ScholarFindings
  • Saenko, K.; Kulis, B.; Fritz, M.; and Darrell, T. 2010. Adapting visual category models to new domains. In ECCV.
    Google ScholarFindings
  • Torralba, A., and Efros, A. A. 2011. Unbiased look at dataset bias. In CVPR.
    Google ScholarFindings
  • Tzeng, E.; Hoffman, J.; Zhang, N.; Saenko, K.; and Darrell, T. 2014. Deep domain confusion: Maximizing for domain invariance. CoRR abs/1412.3474.
    Findings
  • Tzeng, E.; Hoffman, J.; Zhang, N.; Saenko, K.; and Darrell, T. 2015. Simultaneous deep transfer across domains and tasks. In ICCV.
    Google ScholarFindings
  • Wang, X., and Schneider, J. 2014. Flexible transfer learning under support and model shift. In NIPS.
    Google ScholarFindings
  • Yosinski, J.; Clune, J.; Bengio, Y.; and Lipson, H. 2014. How transferable are features in deep neural networks? In NIPS.
    Google ScholarLocate open access versionFindings
  • Zhang, K.; Scholkopf, B.; Muandet, K.; and Wang, Z. 2013. Domain adaptation under target and conditional shift. In ICML.
    Google ScholarFindings
  • Zhong, E.; Fan, W.; Yang, Q.; Verscheure, O.; and Ren, J. 2010. Cross validation framework to choose amongst models and datasets for transfer learning. In ECML/PKDD, 547–562. Springer.
    Google ScholarLocate open access versionFindings
Your rating :
0

 

Tags
Comments