C 2DAN: An Improved Deep Adaptation Network with Domain Confusion and Classifier Adaptation

Han Sun*, Xinyi Chen, Ling Wang, Dong Liang, Ningzhong Liu, Huiyu Zhou

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

1 Downloads (Pure)


Deep neural networks have been successfully applied in domain adaptation which uses the labeled data of source domain to supplement useful information for target domain. Deep Adaptation Network (DAN) is one of these efficient frameworks, it utilizes Multi-Kernel Maximum Mean Discrepancy (MK-MMD) to align the feature distribution in a reproducing kernel Hilbert space. However, DAN does not perform very well in feature level transfer, and the assumption that source and target domain share classifiers is too strict in different adaptation scenarios. In this paper, we further improve the adaptability of DAN by incorporating Domain Confusion (DC) and Classifier Adaptation (CA). To achieve this, we propose a novel domain adaptation method named C2DAN. Our approach first enables Domain Confusion (DC) by using a domain discriminator for adversarial training. For Classifier Adaptation (CA), a residual block is added to the source domain classifier in order to learn the difference between source classifier and target classifier. Beyond validating our framework on the standard domain adaptation dataset office-31, we also introduce and evaluate on the Comprehensive Cars (CompCars) dataset, and the experiment results demonstrate the effectiveness of the proposed framework C2DAN.

Original languageEnglish
Article number3606
Number of pages18
JournalSensors (Switzerland)
Issue number12
Publication statusPublished - 26 Jun 2020
Externally publishedYes

Bibliographical note

Funding Information:
Funding: This work is supported by the Fundamental Research Funds for the Central Universities of China under Grant NS2016091. H. Zhou was supported by UK EPSRC under Grant EP/N011074/1, Royal Society-Newton Advanced Fellowship under Grant NA160342, and European Union’s Horizon 2020 research and innovation program under the Marie-Sklodowska-Curie grant agreement No 720325.

Publisher Copyright:
© MDPI AG. All rights reserved.

Copyright 2020 Elsevier B.V., All rights reserved.


  • Classifier adaptation
  • Domain adaptation
  • Domain confusion
  • MK-MMD
  • Transfer learning
  • Vehicle classification

ASJC Scopus subject areas

  • Analytical Chemistry
  • Biochemistry
  • Atomic and Molecular Physics, and Optics
  • Instrumentation
  • Electrical and Electronic Engineering


Dive into the research topics of 'C 2DAN: An Improved Deep Adaptation Network with Domain Confusion and Classifier Adaptation'. Together they form a unique fingerprint.

Cite this