Abstract
Face super-resolution is an example of super-resolution technique, where it takes one or multiple observed low-resolution images and then converts them to high-resolution image. Learning-based face super-resolution depends on prior information from training database. Most patch-based face super-resolution methods assume the homoscedasticity of the reconstruction error in an objective function and solve it with regularized least squares. In fact, the heteroscedasticity generally exists both in the observed data and in the reconstruction error. To access accurate prior information, we propose a nonlinear adaptive representation (NAR) scheme for hallucinating the individuality of facial images. First, we apply a weighted regularization process to both the reconstruction error and representation coefficients terms to eliminate the heteroscedasticity of the input data. Then, the contextual patches and residual high-frequency components are explored to enrich the prior information. Moreover, a nonlinear extension of the adaptive representation fully utilizes accurate prior information to achieve better reconstruction performance. Experiments on the CAS-PEAL-R1, Webface and LDHF databases show that NAR outperforms some state-of-the-art face super-resolution methods including some deep learning-based approaches.
Similar content being viewed by others
References
Baker S, Kanade T (2000) Hallucinating faces. In: IEEE international conference on automatic face and gesture recognition, p 83
Chang H, Yeung DY, Xiong Y (2004) Super-resolution through neighbor embedding. In: Proceedings of the 2004 IEEE computer society conference on computer vision and pattern recognition, 2004. CVPR 2004, pp I–I
Chen Y, Li F, Fan J (2015) Mining association rules in big data with NGNP. Cluster Comput 18(2):577–585
Dong C, Chen CL, He K, Tang X (2016) Image super-resolution using deep convolutional networks. IEEE Trans Pattern Anal Mach Intell 38(2):295–307
Gao G, Jing XY, Huang P, Zhou Q, Wu S, Yue D (2016) Locality-constrained double low-rank representation for effective face hallucination. IEEE Access 4(99):8775–8786
Gao W, Cao B, Shan S, Chen X, Zhou D, Zhang X, Zhao D (2008) The cas-peal large-scale chinese face database and baseline evaluations. IEEE Trans Syst Man Cybern A Syst Hum 38(1):149–161
Huang Z, Li Q, Zhang T, Sang N, Hong H (2017) Iterative weighted sparse representation for X-ray cardiovascular angiogram image denoising over learned dictionary. IET Image Proc 12(2):254–261
Huang Z, Zhang Y, Li Q, Zhang T, Sang N (2018) Spatially adaptive denoising for X-ray cardiovascular angiogram images. Biomed Signal Process Control 40:131–139
Jiang J, Chen C, Huang K, Cai Z, Hu R (2016) Noise robust position-patch based face super-resolution via Tikhonov regularized neighbor representation. Inf Sci 367(C):354–372
Jiang J, Hu R, Han Z, Lu T, Huang K (2012) Position-patch based face hallucination via locality-constrained representation. In: IEEE international conference on multimedia and expo, pp 212–217
Jiang J, Hu R, Han Z, Wang Z, Lu T, Chen J (2013) Locality-constraint iterative neighbor embedding for face hallucination. In: IEEE international conference on multimedia and expo, pp 1–6
Jung C, Jiao L, Liu B, Gong M (2011) Position-patch based face hallucination using convex optimization. IEEE Signal Process Lett 18(6):367–370
Kang D, Han H, Jain AK, Lee SW (2014) Nighttime face recognition at large standoff: cross-distance and cross-spectral matching. Pattern Recogn 47(12):3750–3766
Kim J, Lee JK, Lee KM (2016) Accurate image super-resolution using very deep convolutional networks. In: Computer vision and pattern recognition, pp 1646–1654
Lan C, Hu R, Han Z et al (2010) A face super-resolution approach using shape semantic mode regularization. In: 2010 IEEE international conference on image processing. IEEE, pp 2021–2024
Lanckriet GRG, Cristianini N, Bartlett P, Ghaoui LE, Jordan MI (2002) Learning the kernel matrix with semidefinite programming. J Mach Learn Res 5(1):27–72
Lipsitz SR, Laird NM, Harrington DP (1994) Weighted least squares analysis of repeated categorical measurements with outcomes subject to nonresponse. Biometrics 50(1):11–24
Liu C, Shum HY, Zhang CS (2001) A two-step approach to hallucinating faces: global parametric model and local nonparametric model. In: Proceedings of the 2001 IEEE computer society conference on computer vision and pattern recognition, 2001. CVPR 2001, vol 1, pp I–192–I–198
Liu J, Wang X, Chen M, Liu S, Shao Z, Zhou X, Liu P (2014) Illumination and contrast balancing for remote sensing images. Remote Sens 6(2):1102–1123
Liu X, Xu Q, Ma J, Jin H, Zhang Y (2014) MsLRR: a unified multiscale low-rank representation for image segmentation. IEEE Trans Image Process 23(5):2159–2167
Lu T, Hu R, Han Z, Jiang J, Zhang Y (2013) From local representation to global face hallucination: a novel super-resolution method by nonnegative feature transformation. In: Visual communications and image processing, pp 1–6
Lu T, Pan L, Wang J, Zhang Y, Wang Z, Xiong Z (2017) AWCR: adaptive and weighted collaborative representations for face super-resolution with context residual-learning. In: Pacific rim conference on multimedia, pp 107–116
Ma J, Zhao J, Jiang J, Zhou H, Guo X (2019) Locality preserving matching. Int J Comput Vis 127(5):512–531
Ma X, Zhang J, Qi C (2010) Hallucinating face by position-patch. Pattern Recogn 43(6):2224–2236
Peng L, Zhang Y, Zhou H, Lu T (2018) A robust method for estimating image geometry with local structure constraint. IEEE Access 6:20734–20747
Romano Y, Elad M (2016) Con-patch: when a patch meets its context. IEEE Trans Image Process 25(9):3967–3978
Roweis ST, Saul LK (2000) Nonlinear dimensionality reduction by locally linear embedding. Science 290(5500):2323–6
Saitoh S (2003) Theory of reproducing kernels. Trans Am Math Soc 68(3):337–404
Shao Z, Chen M, Liu C (2015) Feature matching for illumination variation images. J Electron Imaging 24(3):033011
Shi J, Liu X, Zong Y, Qi C, Zhao G (2018) Hallucinating face image by regularization models in high-resolution feature space. IEEE Trans Image Process PP(99):1–1
Shi W, Caballero J, Huszár F et al (2016) Real-time single image and video super-resolution using an efficient sub-pixel convolutional neural network. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 1874–1883
Tikhonov AN, Arsenin VY (1977) Solution of ill-posed problems. Math Comput 32(144):491–491
Timofte R, De V, Gool LV (2013) Anchored neighborhood regression for fast example-based super-resolution. In: IEEE international conference on computer vision, pp 1920–1927
Timofte R, Smet VD, Gool LV (2014) A+: adjusted anchored neighborhood regression for fast super-resolution. Springer, Berlin
Timofte R, Van Gool L (2012) Weighted collaborative representation and classification of images. In: International conference on pattern recognition, pp 1606–1610
Wang N, Tao D, Gao X, Li X, Li J (2014) A comprehensive survey to face hallucination. Int J Comput Vis 106(1):9–30
Wang X, Tang X (2005) Hallucinating face by eigentransformation. IEEE Press, Piscataway
Wang Z, Bovik AC, Sheikh HR, Simoncelli EP (2004) Image quality assessment: from error visibility to structural similarity. IEEE Trans Image Process 13(4):600–612
Wang Z, Hu R, Jiang J, Han Z, Shao Z (2016) Heteroskedasticity tuned mixed-norm sparse regularization for face hallucination. Multimed Tools Appl 75(24):1–29
Wang Z, Hu R, Wang S, Jiang J (2014) Face hallucination via weighted adaptive sparse regularization. IEEE Trans Circuits Syst Video Technol 24(5):802–813
Wang Z, Liu D, Yang J, Han W, Huang T (2016) Deep networks for image super-resolution with sparse prior. In: IEEE international conference on computer vision, pp 370–378
Waqas J, Yi Z, Zhang L (2013) Collaborative neighbor representation based classification using l 2-minimization approach. Pattern Recogn Lett 34(2):201–208
Yang J, Wright J, Huang T et al (2008) Image super-resolution as sparse representation of raw image patches. In: 2008 IEEE conference on computer vision and pattern recognition. IEEE, pp 1–8
Yang J, Wright J, Huang TS, Ma Y (2010) Image super-resolution via sparse representation. IEEE Trans Image Process 19(11):2861–2873
Yi D, Lei Z, Liao S et al (2014) Learning face representation from scratch. arXiv preprint. arXiv:1411.7923
Yu Y, Tang S, Aizawa K, Aizawa A (2018) Category-based deep CCA for fine-grained venue discovery from multimodal data. IEEE Trans Neural Netw Learn Syst 30(4):1250–1258
Zeng K, Lu T, Liang X et al (2019) Face super-resolution via bilayer contextual representation. Signal Process Image Commun 75:147–157
Zhang L, Zhou WD, Chang PC, Liu J, Yan Z, Wang T, Li FZ (2015) Kernel sparse representation-based classifier. Multimed Tools Appl 74(1):123–137
Zhang Y, Zhang Z, Hu G, Hancock ER (2017) Face image super-resolution via weighted patches regression. In: International conference on pattern recognition, pp 3892–3897
Zhao T, Wang Y, Wang H, Sheng G, Song H (2009) Face recognition method by using large and representative datasets. In: Control and decision conference, 2009. CCDC’09. Chinese
Zhou H, Hu J, Lam KM (2015) Global face reconstruction for face hallucination using orthogonal canonical correlation analysis. In: Asia-Pacific signal and information processing association summit and conference, pp 537–542
Zhou H, Ma J, Yang C, Sun S, Liu R, Zhao J (2016) Nonrigid feature matching for remote sensing images via probabilistic inference with global and local regularizations. IEEE Geosci Remote Sens Lett 13(3):374–378
Zhou H, Ma J, Zhang Y, Yu Z, Ren S, Chen D (2017) Feature guided non-rigid image/surface deformation via moving least squares with manifold regularization. In: 2017 IEEE international conference on multimedia and expo (ICME). IEEE, pp 1063–1068
Zhou L, Wang Z, Luo Y et al (2019) Separability and compactness network for image recognition and superresolution. IEEE Trans Neural Netw Learn Syst 30(11):3275–3286
Zhu S, Liu S, Loy CC et al (2016) Deep cascaded bi-network for face hallucination. In: European conference on computer vision. Springer, Cham, pp 614–630
Acknowledgements
This work is supported by the National Natural Science Foundation of China (61502354, 61501413, 61671332, 41501505), the Natural Science Foundation of Hubei Province of China (2018ZYYD059, 2015CFB451, 2014CFA130, 2012FFA099, 2012FFA134, 2013CF125), Hubei Technology Innovation Project (2019AAA045), Scientific Research Foundation of Wuhan Institute of Technology (K201713). The authors declared that they have no conflicts of interest to this work.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Lu, T., Zeng, K., Qu, S. et al. Face super-resolution via nonlinear adaptive representation. Neural Comput & Applic 32, 11637–11649 (2020). https://doi.org/10.1007/s00521-019-04652-5
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00521-019-04652-5