臺灣博碩士論文加值系統

年齡估測在電腦視覺的領域中是一大挑戰，尤其是透過影像來辨識人的真實年齡，年齡會因身體的狀況或外在因素而造成非常大的誤差值，因此實作上有極大的困難點。年齡估測可以適用在很多領域，例如在娛樂系統、消費系統，分級系統和客群分析上，都能有很好的應用效果。由於現今人民對智慧生活及自動化的要求，時常搭配不同鏡頭再經由各式數位影像處理的演算法不斷擴展來辨識、偵測或是追蹤，因此不管是在學術界與工業界都湧入大量專業領域的人進行開發，致使此領域在各方面都獲得了更好的發揮舞台。在現今也擁有許多相關的年齡估測的演算法，在提取特徵上都是以單特徵搭配固定大小區塊或是在前處理進行特定位置的選取，但是有可能因為這樣忽略了極具代表年齡的特徵，而本論文所提出的方法是透過擴展式的曲率賈柏濾波器(Extended Curvature Gabor Filter，簡稱ECG) 取得臉部的曲線強度，並搭配離散餘旋轉換(Discrete Cosine Transform，簡稱DCT) 來進行特徵降維，再加上我們隨機利用不同的大小、位置及角度產生出非特定區塊並提取該區塊的完整局部二值模式(Completed Local Binary Pattern，簡稱CLBP)及局部方向二值模式(Local Directional Pattern，簡稱LDP)來擷取更有辨識度的特徵。最後利用上述所提取的多特徵組合成特徵向量並透過支持向量機 (Support Vector Machine，簡稱SVM) 中的線性內核演算法來進行年齡估測的迴歸分析，最終預測出年齡。本論文的平均年齡差 (Mean Absolute Error，簡稱 MAE) 可達到4.49歲，比以往的系統能達到更好的效果。

Vision based applications have become a trend in modern world, and among them, age estimation is a useful tool in applications, such as marketing, security, and entertainment. However, age estimation is still quite challenging, because there are many factors, such as environment, mental or physical conditions, would affect the human aging process. Moreover, common make-ups or accessories would occlude important features, such as wrinkles, in pure vision based age estimation systems. There have been many researchers on age estimation, and the proposed system falls in the category of feature-based methods. This thesis proposed a novel method to improve automatic age estimation from human faces. Three types of features extraction algorithms are used, such as Extended Curvature Gabor Filter (ECG), Completed Local Binary Pattern (CLBP), and Local Directional Pattern (LDP). While the ECG is applied to the entire human face, CLBP and LDP are only applied to blocks with randomized scales, positions and orientations. Then, Support Vector Machine (SVM) is used to estimate the age from combined feature vectors. The Mean Absolute Error of the proposed method is 4.49 years old, which is better than existing methods.

摘要 I
ABSTRACT II
致謝 III
LIST OF CONTENTS IV
LIST OF FIGURES VI
LIST OF TABLES VIII
CHAPTER 1 INTRODUCTIONS 1
1.1 Motivation 1
1.2 Contributions 4
1.3 Thesis Organization 5
CHAPTER 2 RELATED WORKS 6
2.1 Facial Model Based Methods 6
2.2 Feature Based Methods 7
2.3 Deep Learning Based Methods 9
CHAPTER 3 PROPOSED METHODS 10
3.1 Preprocessing 11
3.1.1 Face Detection 11
3.1.2 Face Alignment 13
3.2 Feature Extraction 15
3.2.1 Gabor Filters 15
3.2.2 Extended Curvature Gabor Filters 17
3.2.3 Local Binary Pattern 21
3.2.4 Completed Local Binary Pattern 22
3.2.5 Local Directional Pattern 26
3.2.6 Randomized Facial blocks 29
3.3 Dimension Reduction 32
3.3.1 Discrete Cosine Transform 32
3.4 Feature Fusion 35
3.5 Support Vector Machine 37
CHAPTER 4 EXPERIMENTAL RESULTS 38
4.1 Experimental Environment 38
4.2 Age Database 38
4.2.1 FG-NET 39
4.2.2 MORPH 2 40
4.3 Evaluation Performance 41
4.4 Compared existing and analyses 41
4.4.1 Performance evaluation on FG-NET [14] 41
4.4.2 Performance evaluation on MORPH 2 45
4.4.3 Performance evaluation under different ethics and genders 46
4.5 The proposed system under different parameters 48
CHAPTER 5 CONCLUSIONS and Future works 52
5.1 Conclusions 52
5.2 Future Works 53
REFERENCES 54

[1] Customer Demographics: http://www.intumit.com/SmartFace.html, [Online].
[2] Automatic Charging System: https://news.gamme.com.tw/62263, [Online].
[3] Intelligent Vending Machine :https://news.gamme.com.tw/62263, [Online].
[4] T. Wu, P. Turaga, and R. Chellappa, “Age estimation and face verification across aging using landmarks,” IEEE Trans. Information forensics and security, vol. 7, no. 6, pp. 1780-1788, 2012.
[5] X. Geng, K. Smith-Miles, and Z. Hou, “Automatic age estimation based on facial aging patterns,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 29, no.12, pp. 2234-2240, 2007.
[6] G. Guo, Y. Fu, C. R. Dyer, and T. S. Huang, “Image-based human age estimation by manifold learning and locally adjusted robust regression,” IEEE Trans. Image Processing, vol.17, no.7, pp. 1178-1188, 2008.
[7] N. Ramanathan, R. Chellappa, and S. Biswas, "Age progression in human faces: A survey," Visual Languages and Computing, vol. 15, pp. 3349–3361, 2009.
[8] G. Guo, G. Mu, Y. Fu, and T. S. Huang, “Human age estimation using bio inspired features,” Proc. IEEE Conf. Computer Vision and Pattern Recognition, pp. 112-119, 2009.
[9] Y. H. Kwon and D. V. Lobo, “Age classificaton from facial images,” IEEE Conf. Computer Vision and Pattern Recognition, pp. 762-767, 1999.
[10] T. Cootes, G. Edwards and C. Taylor, “Active appearance models,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 23, no. 6, pp. 681-685, Jun 2001.
[11] W. Hwang, X. Huang and K. Noh, “Face recognition system using extended curvature gabor classifier bunch for low-resolution face image,” Computer Vision and Pattern Recognition 2011, pp. 15-22, 2011.
[12] Z. Guo, L. Zhang and D. Zhang, “A completed modeling of local binary pattern operator for texture classification,” IEEE Trans. Image Processing, vol. 19, no. 6, pp. 1657-1663, 2010.
[13] T. Jabid, M. Kabir and O. Chae, “Local directional pattern (LDP) for face recognition,” International Conference On Consumer Electronics 2010, pp. 329-330, 2010.
[14] The FG-NET Aging Database, http://www.fgnet.rsunit.com, 2010 [Online].
[15] K. Ricanek and T. Tesafaye, “MORPH: A longitudinal image database of normal adult age-progression,” in IEEE International Conference on Automatic Face and Gesture Recognition (FG), pp. 341-345, 2006.
[16] A. Günay and V. Nabiyev, “Age estimation based on AAM and 2D-DCT features of facial images,” Internatiomal Journal of Aavanced Computer Science and Applications, vol. 6, no. 2, 2015.
[17] J.K. Pontes, A.S.Britto Jr, C. Fookes, A.L. Koerich, “A flexible hierarchical approach for facial age estimation based on multiple features,” Pattern Recognition, vol. 54, pp. 34-51, June, 2016.
[18] A. Rybintsev, “Age Estimation from a face image in a selected gender-race group based on ranked local binary patterns, ” Complex & Intelligent Systems, vol 3, pp. 93-104, 2017.
[19] D. Karthikeyan and G. Balakrishnan, “Feature vector fusion for image based human age estimation,” International Journal of Computer, Electrical, Automation, Control and Information Engineering, vol. 9, no. 12, 2015.
[20] S.N. Gowda, “Age estimation by LS-SVM regression on facial images,” International Symposium on Visual Computing, pp. 370-379, 2016.
[21] Y. Liang, L. Liu, Y. Xu, Y. Xiang, and B. Zou, “Multi-task GLOH feature Selection for Human age estimation,” International Conference on Image Processing 2011, pp. 565-568, 2011.
[22] D.G. Lowe, “Object recognition from local scale-invariant features,” Proceedings of the Seventh IEEE International Conference on Computer Vision, pp. 20-27, Sept. 1999.
[23] V.T. Selvi and K. Vani, “Age estimation system using MPCA,” International Conference on Recent Trends in Information Technology 2011, pp. 1055-1060, June 3-5, 2011.
[24] H. Takimoto, Y. Mitsukura, M. Fukumi, and N. Akamatsu, “Robust gender and age estimation under varying facial pose,” Electronics and Communications in Japan, vol. 91, no. 7, 2008.
[25] S. E. Choi, Y. J. Lee, S. J. Lee, K. R. Park, and J. Kim, “A comparative study of local feature extraction for age estimation,” Int. Conf. Control, Automation, Robotics and Vision, Dec. 2010.
[26] J. I. Hayashi, H. Koshimizu, and S. Hata, “Age and gender estimation based on facial image analysis,” Int. Conf. Knowledge-Based and Intelligent information and Engineering Systems, pp. 863-869, 2003.
[27] S. Yan, M. Liu, and T. S. Huang, “Extracting age information from local spatially flexible patches,” International Conference on Acoustics, Speech and Signal Processing, 2008.
[28] A. Günay and V. Nabiyev, “Facial age estimation using spatial weber local descriptor,” International Conference on Telecommunications and Signal Processing, pp. 466-469, 2016.
[29] Y. Bengio, P. Lambin, D. Popovici, and H. Larochelle, “Greedy layer-wise training of deep networks, ” Int. Conf. Neural information processing systems, 2006.
[30] G. E. Hinton and R. R. Salakhtdinov, “Reducing the dimensionality of data with neural networks, ” Science, vol. 313, pp. 504-507, 2006.
[31] A. Krizhevsky, I. Sutskever, and G. E. Hinton, “Imagenet classification with deep convolutional neural networks, ” Int. Conf. Neural information processing systems, 2012.
[32] Y. Zhou, H. Chang, K. Barner, P. Spellman, and B. Parvin, “Classification of histology sections via multispectral convolutional sparse coding,” Int. Conf. Computer Vision and Pattern Recognition, pp. 3081-3088, 2014.
[33] Y. Le Cun, B. Boser, J. S. Denker, D. Henderson, R. E. Howard, W. Hubbard, and L. D. Jackel, “Handwritten digit recognition with a back-propagation network, “Int. Conf. Neural information processing systems, 1989.
[34] M. D. Zeiler and R. Fergus, “ Visualizing and understanding convolutional neural networks, “ Europen Conference on Computer Vision, pp. 818-833. 2014.
[35] J.C. Chen, A. Kumar, R. Ranjan, V.M Patel, A. Alavi, and R. Chellappa,"A cascade convolutional neural network for age estimation of unconstrained faces," International Conference on Biometrics Theory, Applications and System, pp. 1-8, 2016.
[36] A. Dosovitskiy, J. T. Springenberg, and T. Brox, “Unsupervised feature learning by augmenting single images,” International Conference on Learning Representations, 2013.
[37] G. Levi and T. Hassncer, “Age and gender classification using convolutional neural networks,” IEEE Conf. Computer Vision and Pattern Recognition, 2015.
[38] P. Viola and M. Jones, “Rapid object detection using a boosted cascade of simple features,” Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. pp. 511-518, 2001.
[39] Open CV: http://opencv.org/, [Online].
[40] Y. Shih and C.F. Chuang, “Automatic extraction of head and face boundaries and facial features,” Information Sciences, vol. 158, pp. 117-130, Jan 2004.
[41] D. Dunn and W.E. Higgins, “Optimal gabor filters for texture segmentation,” IEEE Trans on Image Processing, vol. 4, no. 7, July 1995.
[42] T. Ojala, M. Pietikäinen and D. Harwood, "Comparative study of texture detection and classification algorithms," Pattern Recognition, vol. 29, no. 1, pp. 51-59, 1996.
[43] R. A. Kirsch, “Computer determination of the constituent structure of biological images,” Computers and Biolmedical Research, vol. 4, pp. 315-328, 1971.
[44] N. Ahmed, T. Natarajan and K.R. Rao, “Discrete cosine transform,” IEEE Trans, Computers, vol.C-23, no.1, pp. 90-93, Jan. 1974.
[45] C. Cortes and V. Vapnik, "Support-Vector Networks," Machine Learning, vol. 20, no. 3, p. 273–297, 1995.
[46] A. Lanitis, C. Taylor, and T. Cootes, “Toward automatic simulation of aging effects on face images,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 24, no. 4, pp. 442-455, Apr 2002.
[47] X. Geng, K. Smith-Miles, and Z. Zhou, “Facial age estimation by nonlinear aging patterns subspace,” In Proc. of 16th ACM International Conference on Multimedia, pp. 721-724, 2008.
[48] X. Geng, C. Yin and Z. H. Zhou, “Facial age estimation by learning from label distributions,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 35, no. 10, pp. 2401-2412, Oct 2013.