性別辨識在現代人的生活裡已經有很多的應用。在一般影像處理的方法中，性別辨識主要是依據人臉影像來進行識別。然而在實際拍攝的情況當中，由於複雜背景環境、光源照射、遮蔽以及解析度不佳等問題，常使得臉部偵測失效，造成性別辨識的效果大幅下降。因此，本研究提出了多重屬性(multiple-attributes, MA)辨識方法來作性別辨識。除了對臉部影像辨識以外，還利用各種具有凸顯性別的衣著屬性，作為性別辨識的依據，並再針對穿著屬性位於影像的所在位置，研發出各個部位的自動切割演算法。在辨識的方法中，使用了方向梯度直方圖(histogram of oriented gradient, HOG)演算法抽取特徵，以支持向量機(support vector machine, SVM)分類器進行特徵判別，最後將衣著屬性辨識的結果給予對應的權重值，計算出歸屬於男性的得分數與女性的得分數，依據分數的高低來決定人們的性別。文獻回顧顯示本研究為第一個提出以全身多重屬性的概念來辨識性別，在不同的屬性辨識區塊中，依據屬性所合適的辨識區塊開發出自動切割演算法，以及一套男女屬性權重計分理論。
為了證明加入多重屬性後的性別辨識效果，在實驗中本研究使用了自己建立的全身影像資料庫，並比較了Fisherface的人臉性別辨識方法以及我們所提出的方法。實驗結果顯示加入衣著屬性後在性別辨識的準確率高於單一人臉辨識的準確率，本研究建立了一套強健的智慧型性別辨識系統。

Gender recognition has many applications in modern people’s lives. In the general image processing methods, gender recognition is mainly based on face image recognition. However, in the actual shooting situation, due to complex background environment, light source illumination, shadowing, and poor resolution, face detection often fails, and then the effect of gender recognition is greatly reduced. Therefore, this study proposes a multiple attribute-based (MA) identification method for gender recognition. In addition to facial image recognition, it also uses various clothing attributes, being able to highlight the gender, as the basis for recognition. Moreover, according to the part where the wearing attribute is located, the automatic segmentation algorithm for each part is developed. In the recognition methods, histogram of oriented gradient (HOG) algorithm is used to extract the features, and the features are identified by support vector machine (SVM) classifier. Finally, the corresponding weights are given to the results of the clothing attribute identification, the male score and the female score are calculated, respectively, and then the gender can be determined according to the scores. Literature survey indicates that this study is the first to propose the concept of multiple attributes of the whole body to recognize gender. In different attribute identification blocks, the automatic segmentation algorithms are developed according to the block suitable for the attribute identification, and a set of scoring theories with attribute weights for men and women is proposed.
In order to prove the effect of gender recognition after adding multiple attributes, in the experiment, the self-established whole body image database is used. Fisherface gender identification method and our proposed method are compared. Experimental results show that the accuracy of gender recognition after adding clothing attributes is higher than that of single face recognition. This study has established a robust intelligent gender recognition system.

[1] R. Ravnik and F. Solina,“Interactive and Audience Adaptive Digital Signage Using Real-Time Computer Vision,” vol. 10, no. 2, p. 107, 2013.
[2] M. Y. İşcan, S. R. Loth, C. A. King, D. Shihai, and M. Yoshino,“Sexual dimorphism in the humerus: A comparative analysis of Chinese, Japanese and Thais,” Forensic Science International, vol. 98, no. 1, pp. 17-29, Nov. 1998.
[3] S. Baluja and H. A. Rowley,“Boosting sex identification performance,” vol. 71, no. 1, pp. 111-119, Jan. 2007.
[4] P. Viola and M. Jones,“Rapid object detection using a boosted cascade of simple features,” in Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001, Dec. 2001.
[5] P. Viola and M. J. Jones,“Robust Real-Time Face Detection,” International Journal of Computer Vision, vol. 57, no. 2, pp. 137-154, May. 2004.
[6] P. J. Phillips, H. Wechsler, J. Huang, and P. J. Rauss,“The FERET database and evaluation procedure for face-recognition algorithms,” Image and Vision Computing, vol. 16, no. 5, pp. 295-306, Apr. 1998.
[7] C. Shan,“Learning local binary patterns for gender classification on real-world face images,” Pattern Recognition Letters, vol. 33, no. 4, pp. 431-437, Mar. 2012.
[8] Y. Freund and R. E. Schapire,“A Decision-Theoretic Generalization of On-Line Learning and an Application to Boosting,” Journal of Computer and System Sciences, vol. 55, no. 1, pp. 119-139, Aug. 1997.

[9] R. E. Schapire and Y. Singer,“Improved Boosting Algorithms Using Confidence-rated Predictions,” Machine Learning, vol. 37, no. 3, pp. 297-336, Dec. 1999.
[10] L. Wolf, T. Hassner, and Y. Taigman,“Similarity Scores Based on Background Samples,” in Asian Conference on Computer Vision (ACCV), Berlin, Heidelberg, 2010, pp. 88-97: Springer Berlin Heidelberg.
[11] G. B. Huang, M. Mattar, T. Berg, and E. Learned-Miller,“Labeled Faces in the Wild: A Database forStudying Face Recognition in Unconstrained Environments,” presented at the Workshop on Faces in 'Real-Life' Images: Detection, Alignment, and Recognition, Marseille, France, Oct., 2008.
[12] H. Han, C. Otto, X. Liu, and A. K. Jain,“Demographic Estimation from Face Images: Human vs. Machine Performance,” IEEE Transactions on Pattern Analysis & Machine Intelligence, vol. 37, no. 6, pp. 1148-1161, Jun. 2015.
[13] M. Riesenhuber and T. Poggio,“Hierarchical models of object recognition in cortex,” Nature Neuroscience, Article vol. 2, p. 1019, Nov. 1999.
[14] K. Ricanek and T. Tesafaye,“MORPH: a longitudinal image database of normal adult age-progression,” in 7th International Conference on Automatic Face and Gesture Recognition (FGR06), 2006, pp. 341-345.
[15] J. E. Tapia and C. A. Perez,“Gender Classification Based on Fusion of Different Spatial Scale Features Selected by Mutual Information From Histogram of LBP, Intensity, and Shape,” IEEE Transactions on Information Forensics and Security, vol. 8, no. 3, pp. 488-499, 2013.
[16] J. Bekios-Calfa, J. M. Buenaposada, and L. Baumela,“Revisiting Linear Discriminant Techniques in Gender Recognition,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 33, no. 4, pp. 858-864, Dec. 2011.
[17] E. Eidinger, R. Enbar, and T. Hassner,“Age and Gender Estimation of Unfiltered Faces,” IEEE Transactions on Information Forensics and Security, vol. 9, no. 12, pp. 2170-2179, Sep. 2014.
[18] G. Levi and T. Hassner,“Age and Gender Classification Using Convolutional Neural Networks,” The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, pp. 34-42, 2015.
[19] H.-C. Shih,“Robust gender classification using a precise patch histogram,” Pattern Recognition, vol. 46, no. 2, pp. 519-528, Feb. 2013.
[20] J. Bekios-Calfa, J. M. Buenaposada, and L. Baumela,“Robust gender recognition by exploiting facial attributes dependencies,” Pattern Recognition Letters, vol. 36, pp. 228-234, Jan. 2014.
[21] E. Makinen and R. Raisamo,“Evaluation of Gender Classification Methods with Automatically Detected and Aligned Faces,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 30, no. 3, pp. 541-547, Jan. 2008.
[22] Stegmann, B. Mikkel Bille, K. Ersbøll, and R. Larsen,“FAME - A Flexible Appearance Modelling Environment,” IEEE Transactions on Medical Imaging, vol. 22, no. 10, pp. 1319-1331, 2003.
[23] “Dappei,” URL: https://dappei.com, Internet date: Jul. 2018.
[24] R. Brunelli and T. Poggio,“Face recognition: features versus templates,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 15, no. 10, pp. 1042-1052, Oct. 1993.
[25] E. Meng Joo, W. Shiqian, L. Juwei, and T. Hock Lye,“Face recognition with radial basis function (RBF) neural networks,” IEEE Transactions on Neural Networks, vol. 13, no. 3, pp. 697-710, May. 2002.

[26] A. Anjos, M. M. Chakka, and S. Marcel,“Motion-based counter-measures to photo attacks in face recognition,” IET Biometrics, vol. 3, no. 3, pp. 147-158
[27] Z. Wei-Shi, L. Jian-Huang, and P. C. Yuen,“GA-fisher: a new LDA-based face recognition algorithm with selection of principal components,” IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics), vol. 35, no. 5, pp. 1065-1078, Oct. 2005.
[28] S.-Y. Chiu, C.-C. Chiu, and S. Xu,“A Background Subtraction Algorithm in Complex Environments Based on Category Entropy Analysis,” Advanced Internet of Things for Smart Infrastructure System, vol. 8, no. 6, p. 885, May. 2018.
[29] N. Dalal and B. Triggs,“Histograms of oriented gradients for human detection,” in 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05), 2005, vol. 1, pp. 886-893.
[30] C.-C. Chang and C.-J. Lin,“LIBSVM: A library for support vector machines,”ACM Transactions on Intelligent Systems and Technology (TIST), vol. 2, no. 3, pp. 1-27, Apr. 2011.
[31] P. Rai and P. Khanna,“A gender classification system robust to occlusion using Gabor features based (2D)2PCA,” Journal of Visual Communication and Image Representation, vol. 25, no. 5, pp. 1118-1129, Jul. 2014.
[32] D. Yijun, L. Xiaobo, C. Wujun, and X. Qianzhou,“Gender recognition using fisherfaces and a fuzzy iterative self-organizing technique,” in 2013 10th International Conference on Fuzzy Systems and Knowledge Discovery (FSKD), 2013, pp. 196-200.