Abstract
In reality, however, we always confront such a problem that no target data are achievable, especially when data are multi-modal. Under this situation, the target modality is blind in the training stage, while only the source modality can be obtained. We define such a problem as Missing Modality Problem in transfer learning.
This chapter is reprinted with permission from AAAI and IEEE. “Latent Low-Rank Transfer Subspace Learning for Missing Modality Recognition”. Twenty-Eighth AAAI Conference on Artificial Intelligence, pp. 2921–2927, 2014; “Missing Modality Transfer Learning via Latent Low-Rank Constraint”. IEEE Transactions on Image Processing (TIP), vol. 24, no. 11, pp. 4322–4334, 2015.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsNotes
- 1.
\(X_{\mathrm {S \cdot B}}\)/\(X_{\mathrm {T \cdot B}}\) denote the source/target modalities in the object database B, where \(X_{\mathrm {T \cdot B}}\) is also the missing modality. In addition, \(X_{\mathrm {S \cdot A}}\)/\(X_{\mathrm {T \cdot A}}\) denote the source/target modalities from the auxiliary database A. Note in the illustration, same shape means same dataset and same color means same modality. The whole procedure is: introduce the auxiliary database A with modalities \(X_{\mathrm {S \cdot A}}\) and \(X_{\mathrm {T \cdot A}}\), and then transfer knowledge in two directions: cross-modality transfer (\(\mathrm {T}({\mathrm {M}})\)) and cross-database transfer (\({\mathrm {T}({\mathrm {D}})}\)). In the unified model, P is the shared subspace projection, \(Y_\mathrm {S}\) is pre-learned low-dimensional feature on the source domain \(X_\mathrm {S}\). The source and target domains are coupled by low-rank constraint Z and latent factor L. In addition, two datasets in the source domain are further coupled by Maximum Mean Discrepancy regularizer \(\varOmega (P)= \mathrm {tr}(P^\mathrm {T}\mathscr {M}P)\).
- 2.
- 3.
- 4.
References
Belhumeur PN, Hespanha JP, Kriegman DJ (1997) Eigenfaces vs. fisherfaces: recognition using class specific linear projection. IEEE Trans Pattern Anal Mach Intell 19(7):711–720
Cai J-F, Candès EJ, Shen Z (2010) A singular value thresholding algorithm for matrix completion. SIAM J Optim 20(4):1956–1982
Di H, Jia S, Yunhong W (2012) The buaa-visnir face database instructions. In: IRIP-TR-12-FR-001
Ding Z, Fu Y (2014) Low-rank common subspace for multi-view learning. In: IEEE international conference on data mining. IEEE, pp 110–119
Ding Z, Shao M, Fu Y (2014) Latent low-rank transfer subspace learning for missing modality recognition. In: Proceedings of the 28th AAAI conference on artificial intelligence
Ding Z, Shao M, Fu Y (2015) Missing modality transfer learning via latent low-rank constraint. IEEE Trans Image Process 24(11):4322–4334
Fernando B, Habrard A, Sebban M, Tuytelaars T et al (2013) Unsupervised visual domain adaptation using subspace alignment. In: IEEE international conference on computer vision
Gong B, Shi Y, Sha F, Grauman K (2012) Geodesic flow kernel for unsupervised domain adaptation. In: IEEE conference on computer vision and pattern recognition. pp 2066–2073
Gretton A, Borgwardt KM, Rasch M, Schölkopf B, Smola AJ (2006) A kernel method for the two-sample-problem. In: Advances in neural information processing systems. pp 513–520
He X, Niyogi P (2003) Locality preserving projections. In: Neural information processing systems, vol 16. p 153
Hou C, Nie F, Yi D, Wu Y (2011) Feature selection via joint embedding learning and sparse regression. International joint conference on artificial intelligence 22(1):1324
Jhuo I-H, Liu D, Lee D, Chang S-F (2012) Robust visual domain adaptation with low-rank reconstruction. In: IEEE conference on computer vision and pattern recognition. pp 2168–2175
Jia C, Kong Y, Ding Z, Fu YR (2014) Latent tensor transfer learning for rgb-d action recognition. In: Proceedings of the 22nd ACM international conference on multimedia. ACM, pp 87–96
Klare BF, Jain AK (2013) Heterogeneous face recognition using kernel prototype similarities. IEEE Trans Pattern Anal Mach Intell 35(6):1410–1422
Lewis AS, Malick J (2008) Alternating projections on manifolds. Math Oper Res 33(1):216–234
Li W, Duan L, Xu D, Tsang I (2013) Learning with augmented features for supervised and semi-supervised heterogeneous domain adaptation. IEEE Trans Pattern Anal Mach Intell 36(6):1134–1148
Liu G, Yan S (2012) Latent low-rank representation for subspace segmentation and feature extraction. In: IEEE international conference on computer vision. pp 1615–1622
Liu G, Lin Z, Yan S, Sun J, Yu Y, Ma Y (2013) Robust recovery of subspace structures by low-rank representation. IEEE Trans Pattern Anal Mach Intell 35(1):171–184
Long M, Wang J, Ding G, Pan S, Yu P (2013) Adaptation regularization: a general framework for transfer learning. IEEE Transactions on Knowledge and Data Engineering
Pan SJ, Tsang IW, Kwok JT, Yang Q (2011) Domain adaptation via transfer component analysis. IEEE Trans Neural Netw 22(2):199–210
Shao M, Castillo C, Gu Z, Fu Y (2012) Low-rank transfer subspace learning. In: IEEE 12th international conference on data mining. pp 1104–1109
Shao M, Kit D, Fu Y (2014) Generalized transfer subspace learning through low-rank constraint. Int J Comput Vis 109:1–20
Si S, Tao D, Geng B (2010) Bregman divergence -based regularization for transfer subspace learning. IEEE Trans Knowl Data Eng 22(7):929–942
Turk M, Pentland A (1991) Eigenfaces for recognition. J Cognit Neurosci 3(1):71–86
Wang S, Zhang LY, Pan Q (2012) Semi-coupled dictionary learning with applications in image super-resolution and photo-sketch synthesis. In: IEEE conference on computer vision and pattern recognition. IEEE
Yang J, Yin W, Zhang Y, Wang Y (2009) A fast algorithm for edge-preserving variational multichannel image restoration. SIAM J Imaging Sci 2(2):569–592
Zhang W, Wang X, Tang X (2011) Coupled information-theoretic encoding for face photo-sketch recognition. In: IEEE conference on computer vision and pattern recognition. IEEE, pp 513–520
Zhou T, Tao D (2011) Godec: randomized low-rank & sparse matrix decomposition in noisy case. In: Proceedings of the 28th international conference on machine learning. pp 33–40
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
Copyright information
© 2019 Springer Nature Switzerland AG
About this chapter
Cite this chapter
Ding, Z., Zhao, H., Fu, Y. (2019). Missing Modality Transfer Learning. In: Learning Representation for Multi-View Data Analysis. Advanced Information and Knowledge Processing. Springer, Cham. https://doi.org/10.1007/978-3-030-00734-8_7
Download citation
DOI: https://doi.org/10.1007/978-3-030-00734-8_7
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-00733-1
Online ISBN: 978-3-030-00734-8
eBook Packages: Computer ScienceComputer Science (R0)