Abstract
In this paper, we present a novel approach to address the problem of eye shape characterization via curve representation. Firstly, a directional triangle-area curve representation method (DTAR) is presented for this aim. Equipped with DTAR, the shape similarities between two eyes can be measured by the similarities between two corresponding DTAR curves. Secondly, in order to exploit the underlying information of eye shapes, a curve clustering algorithm is utilized to automatically discover a set of eye shape prototypes. Consequently, a semantics extraction method for eye shapes is proposed in terms of seven reference eye shapes. Finally, in order to validate the consistency of the clustering results and the extracted semantics, extensive experiments on AR and BU4DFE databases are designed and conducted, and all the results demonstrate the effectiveness of the proposed DTAR curve representation and the semantics extraction method.
Similar content being viewed by others
References
Ahonen T, Hadid A, Pietikäinen M (2004) Face recognition with local binary patterns. In: Computer vision-eccv 2004. Springer, pp 469–481
Ahonen T, Rahtu E, Ojansivu V, Heikkilä J (2008) Recognition of blurred faces using local phase quantization. In: 2008. ICPR 2008. 19th international conference on Pattern recognition. IEEE, pp 1–4
Alajlan N, El Rube I, Kamel M, Freeman G (2007) Shape retrieval using triangle-area representation and dynamic space warping. Pattern Recogn 40 (7):1911–1920
An L, Zou C, Zhang L, Denney B (2016) Scalable attribute-driven face image retrieval. Neurocomputing 172:215–224
Baiduwenku: Eye shape classification. http://wenku.baidu.com
Berretti S, Del Bimbo A, Pala P (2000) Retrieval by shape similarity with perceptual distance and effective indexing. IEEE Trans Multimed 2(4):225–239
Borràs A., Lladós J (2005) Object image retrieval by shape content in complex scenes using geometric constraints. In: Pattern recognition and image analysis. Springer, pp 325–332
Cao X, Wei Y, Wen F, Sun J (2014) Face alignment by explicit shape regression. Int J Comput Vis 107(2):177–190
Castellano G, Fanelli AM, Torsello MA (2011) A fuzzy set approach for shape-based image annotation. In: Fuzzy logic and applications. Springer, pp 236–243
Chen X, An L, Bhanu B (2013) Improving large-scale face image retrieval using multi-level features. In: 2013 20th IEEE international conference on Image processing (ICIP). IEEE, pp 4367–4371
Conilione P, Wang D (2012) Fuzzy approach for semantic face image retrieval. Comput J 55(9):1130–1145
Davies G, Ellis H, Shepherd J (1981) Perceiving and remembering faces. Academic Press, New York
DeSarbo WS, Cron WL (1988) A maximum likelihood methodology for clusterwise linear regression. J Classif 5(2):249–282
Ding C, Tao D (2015) A comprehensive survey on pose-invariant face recognition. ACM Trans Intell Syst Technol 7(3):37
El Rube I, Alajlan N, Kamel M, Ahmed M, Freeman G (2005) Efficient multiscale shape-based representation and retrieval. In: Image analysis and recognition. Springer, pp 415–422
Everitt B, Hand D (1981) Finite mixture distributions. Chapman and Hall, Ltd
Gaffney S, Smyth P (1999) Trajectory clustering with mixtures of regression models. In: Proceedings of the fifth ACM SIGKDD international conference on Knowledge discovery and data mining. ACM, pp 63–72
Gaffney SJ (2004) Probabilistic curve-aligned clustering and prediction with regression mixture models. Ph.D. thesis, University of California Irvine
Gaffney S (2005) The curve clustering toolbox (cctoolbox). http://www.datalab.uci.edu/resources/CCT
Guler R, Tari S, Unal G (2016) Landmarks inside the shape: Shape matching using image descriptors. Pattern Recogn 49:79–88
Han CC, Liao HYM, Yu GJ, Chen L (2000) Fast face detection via morphology-based pre-processing. Pattern Recogn 33(10):1701–1712
Hannachi A, O’Neill A (2001) Atmospheric multiple equilibria and non-gaussian behaviour in model simulations. Q J R Meteorol Soc 127(573):939–958
Happy S, Routray A (2015) Automatic facial expression recognition using features of salient facial patches. IEEE Trans Affect Comput 6(1):1–12
Hernandez-Matamoros A, Bonarini A, Escamilla-Hernandez E, Nakano-Miyatake M, Perez-Meana H (2016) Facial expression recognition with automatic segmentation of face regions using a fuzzy based classification approach. Knowl-Based Syst 110:1–14
Hsu RL, Jain AK (2002) Semantic face matching. In: 2002. ICME’02. Proceedings. 2002 IEEE international conference on Multimedia and expo. IEEE, vol 2, pp 145–148
Kaur H, Jyoti K (2013) Survey of techniques of high level semantic based image retrieval. IJRCCT 2(1):015–019
Kim H, Park J, Kim HW, Hwang E, Rho S (2019) Robust facial landmark extraction scheme using multiple convolutional neural networks. Multimedi Tools Appl 78(3):3221–3238
Klassen E, Srivastava A, Mio W, Joshi SH (2004) Analysis of planar shapes using geodesic paths on shape spaces. IEEE Trans Pattern Anal Mach Intell 26 (3):372–383
Kontschieder P, Donoser M, Bischof H (2010) Beyond pairwise shape similarity analysis. In: Computer vision–ACCV 2009. Springer, pp 655–666
Kovac J, Peer P, Solina F (2003) Human skin color clustering for face detection. IEEE, vol 2
Krishnan T, McLachlan G (1997) The em algorithm and extensions. Wiley 1 (997):58–60
Kumar N, Berg AC, Belhumeur PN, Nayar SK (2011) Describable visual attributes for face verification and image search. IEEE Trans Pattern Anal Mach Intell 33(10):1962–1977
Liang X, Zhuang Q, Cao N, Zhang J (2009) Shape modeling and clustering of white matter fiber tracts using fourier descriptors. In: 2009. CIBCB’09. IEEE symposium on Computational intelligence in bioinformatics and computational biology. IEEE, pp 292–297
Liang A, Liu W, Li L, Farid MR, Le V (2014) Accurate facial landmarks detection for frontal faces with extended tree-structured models. In: 2014 22nd international conference on Pattern recognition (ICPR). IEEE, pp 538–543
Liu G, Liu S, Muhammad K, Sangaiah A, Doctor F (2018) Object tracking in vary lighting conditions for fog based intelligent surveillance of public spaces. IEEE Access 6(1):29,283–29,296
Liu S, Liu G, Zhou H (2019) A robust parallel object tracking method for illumination variations. Mob Netw Appl 24(1):5–7
Lumini A, Nanni L (2006) A clustering method for automatic biometric template selection. Pattern Recogn 39(3):495–497
Ma WY, Manjunath BS (1999) Netra: a toolbox for navigating large image databases. Multimed Syst 7(3):184–198
Martinez AM, Benavente R (1998) The AR face database. CVC Technical Report #24, June
McLachlan G, Peel D (2004) Finite mixture models. Wiley, New York
Mclachlan GJ, Basford KE (1988) Mixture models: Inference and application to clustering. Marcel Dekker, New York
Nawaf YA, Nixon MS, Hare JS (2018) Semantic face signatures: Recognizing and retrieving faces by verbal descriptions. IEEE Trans Inf Forensic Secur 13(3):706–716
Pedrycz W, Amato A, Lecce VD, Piuri V (2008) Fuzzy clustering with partial supervision in organization and classification of digital images. IEEE Trans Fuzzy Syst 16(4):1008–1026
Peng K, Chen L, Ruan S, Kukharev G (2005) A robust agorithm for eye detection on gray intensity face without spectacles. J Comput Sci Technol 5(3):127–132
Rajpoot NM, Arif M (2008) Unsupervised shape clustering using diffusion map. Ann BMVA 2008(5):17
Ren Y, Li Q, Liu W, Li L (2016) Semantic facial descriptor extraction via axiomatic fuzzy set. Neurocomputing 171:1462–1474
Sezavar A, Farsi H, Mohamadzadeh S (2019) Content-based image retrieval by combining convolutional neural networks and sparse representation. Multimedia Tools and Applications. https://doi.org/10.1007/s11042-019-7321-1
Shahabi C, Safar M (2007) An experimental study of alternative shape-based image retrieval techniques. Multimed Tools Appl 32(1):29–48
Shanmugavadivu P, Sumathy P, Vadivel A (2016) Fosir: Fuzzy-object-shape for image retrieval applications. Neurocomputing 171:719–735
Sheng W, Howells G, Fairhurst M, Deravi F (2008) Template-free biometric-key generation by means of fuzzy genetic clustering. IEEE Trans Inf Forensic Secur 3(2):183–191
Sinha P, Balas B, Ostrovsky Y, Russell R (2007) Face recognition by humans: 20 results all computer vision researchers should know about. Proc IEEE 94 (11):1948–1962
Smyth P, Ide K, Ghil M (1999) Multiple regimes in northern hemisphere height fields via mixture model clustering. J Atmos Sci 56(21):3704–3723
Song F, Tan X, Chen S, Zhou ZH (2013) A literature survey on robust and efficient eye localization in real-life scenarios. Pattern Recogn 46(12):3157–3173
Song X, Jiang S, Herranz L (2017) Multi-scale multi-feature context modeling for scene recognition in the semantic manifold. IEEE Trans Image Process PP(99):2721–2735
Tafaj E, Kasneci G, Rosenstiel W, Bogdan M (2012) Bayesian online clustering of eye movement data. In: Proceedings of the Symposium on Eye Tracking Research and Applications. ACM, pp 285–288
Thang ND, Rasheed T, Lee YK, Lee S, Kim TS (2011) Content-based facial image retrieval using constrained independent component analysis. Inf Sci 181 (15):3162–3174
Tzortzis G, Likas A, Tzortzis G (2014) The minmax k-means clustering algorithm. Pattern Recogn 47(7):2505–2516
Wang S, Jiang S (2015) Instre: a new benchmark for instance-level object retrieval and recognition. Acm Trans Multimed Comput Commun Appl 11(3):1–21
Wang C, Zhang Q, Duan X, Gan J (2018) Multi-ethnical chinese facial characterization and analysis. Multimed Tools Appl 77(4):1–19
Wu Z, Ke Q, Sun J, Shum HY (2011) Scalable face image retrieval with identity-based quantization and multireference reranking. IEEE Trans Pattern Anal Mach Intell 33(10):1991–2001
Xiong X, De la Torre F (2013) Supervised descent method and its applications to face alignment. In: 2013 IEEE conference on Computer vision and pattern recognition (CVPR). IEEE, pp 532–539
Yin PY (2008) Pattern Recognition Techniques, Technology and Applications. InTech
Zhang X, Yin L, Cohn JF, Canavan S, Reale M, Horowitz A, Liu P (2013) A high-resolution spontaneous 3d dynamic facial expression database. In: 2013 10th IEEE international conference and workshops on Automatic face and gesture recognition (FG). IEEE, pp 1–6
Zhang L, Mistry K, Neoh SC, Lim CP (2016) Intelligent facial emotion recognition using moth-firefly optimization. Knowl-Based Syst 111:248–267
Zhu X, Ramanan D (2012) Face detection, pose estimation, and landmark localization in the wild. In: 2012 IEEE conference on Computer vision and pattern recognition (CVPR). IEEE, pp 2879–2886
Acknowledgments
We appreciate the efforts of two reviewers and Associate Editor for their critical comments on this manuscript and make the quality of this paper improved significantly. This work is supported by State Scholarship Fund, Natural Science Foundations of China under Grant No.61602321 and No.61363066, Aviation Science Foundation with No. 2017ZC54007 and 2017ZA54006, Science Fund Project of Liaoning Province Education Department with No.L201614, Natural Science Fund Project of Liaoning Province with No.20170540694, No.20170540692 and No.2015020069, and the Doctor Startup Foundation of Shenyang Aerospace University 13YB11.
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.
Appendices
Appendix
A. Clustering by density estimation
Assume that the d-dimensional vector yi is modeled by the mixture density
in which αk is the k-th mixture weight and pk is the k-th component density with parameter vector 𝜃k. The mixture weights αk sum to one and are nonnegative.
The component densities pk(⋅) model individual specific sub-regions of density, while the mixture p(⋅) summarizes all of these sub-regions according to the mixture weights αk.
The likelihood of a data set Y = y1,…,yn is any function of Θ that is proportional to the probability of the data p(Y |Θ). The log-likelihood is the log of the likelihood and takes the form
assuming the yi are i.i.d. (independently and identically distributed). ML estimates of the parameter vector Θ correspond to values of Θ that maximize (11).
EM is an approximate root-finding procedure that is used to seek the root of the likelihood equation–it iteratively searches for a set of parameters \(\hat {\Theta }\) that maximize the probability of the observed data. EM provides an efficient framework for parameter estimation in the mixture context.
A particular advantage of the probabilistic approach is that the component PDFs can be defined on non-vector data. For example, suppose that yi is now a sequence of curve measurements of length ni, observed at the ni time points in xi. We can define a cluster-specific conditional probabilistic model pk(yi|xi,𝜃k) that relates yi to xi.
In fact, the overall density of yi (now given xi) is a mixture of the component PDFs. In other words, the density takes the form
This conditional mixture density is now defined on curves in this paper. This curve density can be used in place of the mixture density in (10) to establish a model-based clustering procedure for curves. The generative model for the curve mixture is as follows:
-
Assign the i-th curve to cluster k with probability αk,
-
Generate the mean curve for cluster k according to the component density model pk,
-
Define the i-th curve yi to be equal to the mean curve plus some randomly generated noise (e.g., add a Gaussian error term).
B. Polynomial regression mixtures
In the following, a curve clustering methodology based on polynomial regression mix- ture models (PRM) is described. PRMs employ polynomial regression models with Gaussian error terms as the component PDFs. The inclusion of these regression models into the model-based curve clustering framework outlined above leads to an efficient EM learning algorithm for curve clustering.
(a) Model definition
Suppose we have a set Y of n curves as y1,…,yi,…,yn. Each curve has (a possibly unique) length of ni with measurements observed at the points (or times) in xi.A p-th order polynomial regression relationship between yi and xi is assumed with an additive Gaussian error term (a common assumption in the presence of multiple exogenous, unexplained effects).
The regression of yi on xi can be summarized with the following equation:
where the ni × p regression matrix Xi is the Vandermonde matrix evaluated at xi, and β is the p-vector of regression coefficients. The p-th order Vandermonde matrix evaluated at xi is equal to
This regression equation, along with the error model, defines the conditional PDF of yi given xi as \( \mathcal {N}(\mathrm {\mathbf {y}}_{i}|\mathrm {\mathbf {X}}_{i}\beta ,\sigma ^{2} \textit {\textit {I}})\). This PDF represents a probabilistic curve model that naturally allows for curves of variable length with unique measurement intervals and missing observations. Furthermore, the polynomial fit also takes advantage of smoothness information existed in the data.
This PDF can be incorporated into a mixture density by adding dependence of this PDF on k. In notation, this dependence is added in the form of subscripts on the parameters as \(\{\beta _{k},{\sigma ^{2}_{k}}\}\). The incorporation of these cluster-dependent PDFs into the conditional mixture density in (12) results in the definition of the PRMs as
The log-likelihood takes the sum over all n curves of this conditional density with the following form
This function is used to calculate the out-of-sample test log-likelihood scores by substituting in an unseen dataset Y′ for Y. This model can now be used to derive the EM learning algorithm for curve clustering with PRMs.
(b) EM algorithm for PRMs
For the sake of notational simplicity, we also assume that every PDF is implicitly conditioned on a set of parameters (e.g., Θ or 𝜃k), and thus we leave out the explicit dependence on the parameter vector in our notation.
We begin by letting zi give the cluster membership for curve i, and we write the joint density of yi and zi as
The cluster memberships {zi} are regarded as being hidden. The hidden-data density then becomes the posterior p(zi|yi,xi). The complete-data log-likelihood function \(\mathcal {L}_{c}\) can be calculated by taking the sum over all n curves of the log joint density in (16):
E-step
In the E-step, we calculate the posterior p(zi|yi,xi) which gives the membership probability that the i-th curve was generated from cluster zi.The membership probability takes the form
The posterior expectation of \(\mathcal {L}_{c}\) in (17) is then taken with respect to the posterior above to get the Q-function as follows:
M-step
In the M-step, we maximize Q with respect to the parameters \(\{\beta _{k},{\sigma ^{2}_{k}},\alpha _{k}\}\). The solutions are straightforward and are given as
and
Initialization is carried out by randomly sampling values for the membership probabilities and then beginning the iterations with the M-step. Convergence is detected when the ratio of the incremental improvement in log-likelihood to the initial incremental improvement during the second iteration drops below a threshold (e.g., 1 × E− 6 in this paper).
Rights and permissions
About this article
Cite this article
Ren, Y., Li, Q., Liu, W. et al. Semantics characterization for eye shapes based on directional triangle-area curve clustering. Multimed Tools Appl 78, 25373–25406 (2019). https://doi.org/10.1007/s11042-019-7659-4
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11042-019-7659-4