Skip to main content
SpringerLink
Log in

Efficient dynamic domain adaptation on deep CNN

  • Published:
Multimedia Tools and Applications Aims and scope Submit manuscript

Abstract

Domain adaptation is widely-used in deep neural networks to address the problem of data distribution shift. Most of the deep CNN models use the Maximum Mean Discrepancy(MMD) to measure the distribution difference between the source and task domains, which have achieved great success on transfer learning tasks. However, these conventional domain adaptation methods have limited use in dynamic knowledge adaptation due to the constant transfer coefficient for all adaptation layers. In this paper, we propose an efficient dynamic domain adaptation method on deep CNN models, which transfers knowledge dynamically according to different layers and training extent. Specifically, we first present a deep understanding of how transferable each layer in various deep CNN models is, including the different VGG and AlexNet architectures. Then, a dynamic transfer coefficient is proposed based on the deep understanding. By doing this, our paper gives guidance on how to choose transferring layers and adaptation coefficients aptly instead of using empirical constant transfer parameters in conventional methods. Extensive experiments conducted on standard benchmark datasets demonstrate that the proposed method achieves the state-of-the-art results by using dynamic domain adaptation parameters compared with conventional methods.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11
Fig. 12
Fig. 13

Similar content being viewed by others

References

  1. Chen C, Chen Z, Jiang B, et al (2018) Joint domain alignment and discriminative feature learning for unsupervised deep domain adaptation[J]

  2. Dai W, Yang Q, Xue GR, et al (2007) Boosting for transfer learning[C]. International Conference on Machine Learning

  3. Ding H, Zhou H, Zhou SK, et al (2017) A deep cascade network for unaligned face attribute classification[J]

  4. Ding Z, Shao M, Fu Y (2018) Incomplete multisource transfer learning[J]. IEEE Trans Neural Netw Learning Sys 29(2):310–323

    Article  MathSciNet  Google Scholar 

  5. Donahue J, Jia Y, Vinyals O, et al. (2013) DeCAF: a deep convolutional activation feature for generic visual recognition[J]

  6. Fernando B, Habrard A, Sebban M, et al (2013) Unsupervised visual domain adaptation using subspace alignment[C]. International Conference on Computer Vision (ICCV). IEEE

  7. Gong B, Shi Y, Sha F, et al (2012) Geodesic flow kernel for unsupervised domain adaptation[C]. In: 2012 IEEE conference on IEEE computer vision and pattern recognition (CVPR). IEEE

  8. Gopalan R, Li R, Chellappa R (2011) Domain adaptation for object recognition: an unsupervised approach[C]. IEEE International Conference on Computer Vision

  9. Guo Y, Shi H, et al (2018) SpotTune: transfer learning through adaptive fine-tuning[J]

  10. He K, Zhang X, Ren S, et al (2015) Deep residual learning for image recognition[J]

  11. Krizhevsky A, Sutskever I, Hinton G (2012) ImageNet classification with deep convolutional neural networks[C]. NIPS Curran Associates Inc

  12. Li J, Lu K, Huang Z, et al (2018) Heterogeneous domain adaptation through progressive alignment[J]. IEEE Trans Neural Netw Learn Sys 30(5):1381–1391

    Article  MathSciNet  Google Scholar 

  13. Li J, Lu K, Huang Z, et al (2018) Transfer independently together: a generalized framework for domain adaptation[J]. IEEE Trans Cybern 49(6):2144–2155

    Article  Google Scholar 

  14. Li S, Song S, Wu C (2018) Layer-wise domain correction for unsupervised domain adaptation[J]. Frontiers Inform Technol Electron Eng 19(1):91–103

    Article  Google Scholar 

  15. Liu G, Qiu Z, Qu H, et al (2015) Computing k shortest paths using modified pulse-coupled neural network[J]. Neurocomputing 149:1162–1176

    Article  Google Scholar 

  16. Long M, Cao Y, Wang J, et al (2015) Learning transferable features with deep adaptation networks[J]

  17. Long M, Cao Z, Wang J, et al (2017) Conditional adversarial domain adaptation[J]

  18. Long M, Wang J, Ding G, et al (2014) Transfer feature learning with joint distribution adaptation[C]. IEEE International Conference on Computer Vision

  19. Long M, Zhu H, Wang J, et al (2016) Deep transfer learning with joint adaptation networks[J]

  20. Lu H, Li Y, Uemura T, et al (2018) Low illumination underwater light field images reconstruction using deep convolutional neural networks[J]. Futur Gener Comput Syst 82:142–148

    Article  Google Scholar 

  21. Najafabadi MM, Villanustre F, Khoshgoftaar TM, et al (2015) Deep learning applications and challenges in big data analytics[J]. J Big Data 2(1):1

    Article  Google Scholar 

  22. Pan SJ, Yang Q (2010) A survey on transfer learning[J]. IEEE Trans Knowl Data Eng 22(10):1345–1359

    Article  Google Scholar 

  23. Pan SJ, Tsang IW, Kwok JT, et al (2010) Domain adaptation via transfer component analysis[J]. IEEE Trans Neural Netw 22(2):199–210

    Article  Google Scholar 

  24. Qiu S, Deng W (2018) Deep local descriptors with domain adaptation[J]

  25. Russakovsky O, Deng J, Su H, et al (2014) ImageNet large scale visual recognition challenge[J]. Int J Comput Vis 115(3):211–252

    Article  MathSciNet  Google Scholar 

  26. Saenko K, Kulis B, Fritz M, et al (2010) Adapting visual category models to new domains[C]. European Conference on Computer Vision

  27. Saenko K, Kulis B, Fritz M, et al (2010) Adapting visual category models to new domains[C]. European Conference on Computer Vision

  28. Scholkopf B, Platt J, Hofmann T (2006) Correcting sample selection bias by unlabeled data[C]. International Conference on Neural Information Processing Systems

  29. Sun B, Saenko K (2016) Deep CORAL: correlation alignment for deep domain adaptation[M]. Computer Vision - ECCV 2016 Workshops

  30. Tzeng E, Hoffman J, Darrell T, et al (2015) Simultaneous deep transfer across domains and tasks[J]

  31. Tzeng E, Hoffman J, Zhang N, et al (2014) Deep domain confusion: maximizing for domain invariance[J]. Comput Sci

  32. Wen J, Xu Y, Liu H (2018) Incomplete multiview spectral clustering with adaptive graph learning[J]. IEEE transactions on cybernetics, 2018

  33. Xie X, Liu G, Cai Q, et al (2019) The maximum points-based supervised learning rule for spiking neural networks[J]. Soft Computing 23(20):10187–10198

    Article  Google Scholar 

  34. Xie X, Qu H, Liu G, et al (2017) Efficient training of supervised spiking neural networks via the normalized perceptron based learning rule[J]. Neurocomputing 241:152–163

    Article  Google Scholar 

  35. Xie X, Qu H, Yi Z, et al (2017) Efficient training of supervised spiking neural network via accurate synaptic-efficiency adjustment method[J]. IEEE Trans Neural Netw Learning Sys 28(6):1411–1424

    Article  Google Scholar 

  36. Yosinski J, Clune J, Bengio Y, et al (2014) How transferable are features in deep neural networks?[J]. Eprint Arxiv 27:3320–3328

    Google Scholar 

  37. Zhang Y, Zhang Y, Yang Q (2018) Parameter transfer unit for deep neural networks[J]

  38. Zhang Z, Liu L, Shen F et al (2018) Binary multi-view clustering[J]. IEEE Trans Pattern Anal Mach Intell PP(99):1–1

    Google Scholar 

  39. Zhu L, Huang Z, Li Z, et al (2018) Exploring auxiliary context: discrete semantic transfer hashing for scalable image retrieval[J]. IEEE transactions on neural networks and learning systems 29(11):5264–5276

    Article  MathSciNet  Google Scholar 

  40. Zhu L, Huang Z, Liu X, et al (2017) Discrete multi-modal hashing with canonical views for robust mobile landmark search[J]. IEEE Trans Multimed 19(9):2066–2079

    Article  Google Scholar 

  41. Zhu L, Shen J, Xie L, et al (2017) Unsupervised visual hashing with semantic assistant for content-based image retrieval[J]. IEEE Trans Knowl Data Eng 29(2):472–486

    Article  Google Scholar 

Download references

Acknowledgments

This work was supported by the National Natural Science Foundation of China (Grant No: 61806040, 61771098 and 61573081), the China Postdoctoral Science Foundation (Grant No: 2018M633348), and the fund from the Department of Science and Technology of Sichuan Province (Grant No: 2017GFW0128, 18ZDYF2268, 2018JY0578 and 2017JY0007).

Author information

Authors and Affiliations

  1. School of Computer Science, Zhongshan Institute, University of Electronic Science and Technology of China, Zhongshan, 528402, China

    Zeheng Yang & Guisong Liu

  2. School of Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu, 611731, China

    Zeheng Yang, Guisong Liu & Xiurui Xie

  3. Institute for Infocomm Research, Agency for Science, Technology and Research, Singapore, 138632, Singapore

    Xiurui Xie

  4. School of Electrical and Computer Engineering, National University of Singapore, Singapore, 119077, Singapore

    Qing Cai

Authors
  1. Zeheng Yang
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Guisong Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Xiurui Xie
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Qing Cai
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Xiurui Xie.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Yang, Z., Liu, G., Xie, X. et al. Efficient dynamic domain adaptation on deep CNN. Multimed Tools Appl 79, 33853–33873 (2020). https://doi.org/10.1007/s11042-019-08584-z

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11042-019-08584-z

Keywords

Search

Navigation