Abstract
This chapter addresses the problem of transfer learning by unsupervised domain adaptation . We introduce a pipeline which is designed for the case where the joint distribution of samples and labels \(P(\mathbf {X}^{src},\mathbf {Y}^{src})\) in the source domain is assumed to be different, but related to that of a target domain \(P(\mathbf {X}^{ trg },\mathbf {Y}^{ trg })\), but labels \(\mathbf {Y}^{ trg }\) are not available for the target set. This is a problem of Transductive Transfer Learning. In contrast to other methodologies in this book, our method combines steps that adapt both the marginal and the conditional distributions of the data.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
This chapter is an amalgamation of the works published in [152–154] with additional analysis taking into account the works of other authors which were developed concurrently to our work.
- 2.
Note however that in Fig. 6.2b a 2D view of feature space was generated using PCA and only two out of ten classes of digits in MNIST/USPS dataset are shown, while the MMD computation was done in a higher dimensional space with samples from all ten classes. For these reasons it may not be easy to see that the means of the source and target samples became closer after MMD.
- 3.
- 4.
Table 6.3 shows these two measures computed on all datasets, discussed later.
- 5.
The measures were judged as high or low based on a subset of values observed in Table 6.3.
Acknowledgements
N. FarajiDavar and T. deCampos were both at the CVSSP, University of Surrey when the experiments reported in this chapter were developed. This work was supported by the Engineering and Physical Sciences Research Council (EPSRC) Grant number EP/K014307/2 and the MOD University Defence Research Collaboration in Signal Processing.
Author information
Authors and Affiliations
Corresponding authors
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2017 Springer International Publishing AG
About this chapter
Cite this chapter
Farajidavar, N., de Campos, T., Kittler, J. (2017). Adaptive Transductive Transfer Machines: A Pipeline for Unsupervised Domain Adaptation. In: Csurka, G. (eds) Domain Adaptation in Computer Vision Applications. Advances in Computer Vision and Pattern Recognition. Springer, Cham. https://doi.org/10.1007/978-3-319-58347-1_6
Download citation
DOI: https://doi.org/10.1007/978-3-319-58347-1_6
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-58346-4
Online ISBN: 978-3-319-58347-1
eBook Packages: Computer ScienceComputer Science (R0)