基于迁移学习与支持向量机的服装舒适度评估

doi:10.13475/j.fzxb.20191101007

摘要/Abstract

摘要：

针对传统服装舒适度评估需要直接试穿服装导致的舒适度评估精确度不高和评估过程耗时的问题,提出一种从试穿服装数据库学习服装舒适度评估模型的方法,可以快速准确地评估服装舒适度。首先,采集试衣模特尺寸和试穿样板图,并利用迁移学习改善试穿样板图构建试穿服装数据库,同时提出基于虚拟试衣技术的舒适度标签获取方法,为数据库中对应的试穿样板图添加舒适度标签;然后,提取试穿样板图的局部二值模式为服装样板特征,并融合试衣模特尺寸数据形成服装试穿特征向量;最后,提取试穿服装数据库的融合特征训练支持向量机,得到服装舒适度评估模型。实验结果表明,该方法的准确率和系统时间分别为0.8344和12 s,具有较高的精确度和效率。

关键词: 服装舒适度评估, 迁移学习, 虚拟试衣, 特征融合, 支持向量机

Abstract:

The traditional methods for clothing comfort evaluation is carried out through the try-on effect of the garment, which requires much time but with low evaluation accuracy. This paper presented a clothing comfort evaluation model learning from clothing patterns based on the transfer learning and support vector machine fast and accurately. The sizes of mannequins and the graphs of garment patterns were firstly collected, and the graphs of garment patterns were improved by using transfer learning to create garment pattern database. Then, a comfort label acquisition method was presented based on Virtual Try-On, adding comfort label to the corresponding graph of garment pattern. Following that, local binary pattern was extracted from the graph of garment pattern, and it was combined with the sizes of the corresponding mannequins to form clothing comfort feature vector. Finally, the clothing comfort feature vectors of garment pattern database were extracted to train the support vector machine. This exercise shows that the accuracy and average time to evaluate clothing comfort using this method are 0.834 and 12 s respectively, representing satisfactory accuracy and efficiency.

Key words: clothing comfort evaluation, transfer learning, virtual try-on, feature fusion, support vector machine

中图分类号:

TS941

夏海浜, 黄鸿云, 丁佐华. 基于迁移学习与支持向量机的服装舒适度评估[J]. 纺织学报, 2020, 41(06): 125-131.

XIA Haibang, HUANG Hongyun, DING Zuohua. Clothing comfort evaluation based on transfer learning and support vector machine[J]. Journal of Textile Research, 2020, 41(06): 125-131.

图/表 10

图1

表1

图2

图3

图4

表2

图5

图6

表3

表4

参考文献 19

[1]	CHEN Y, ZENG X, HAPPIETTE M, et al. Fuzzy applications in industrial engineering[M]. Berlin: Springer, 2006: 367-379.
[2]	LIU K X, ZENG X Y, BRUNIAUX P, et al. Fit evaluation of virtual garment try-on by learning from digital pressure data[J]. Knowledge-Based Systems, 2017,133:174-182. doi: 10.1016/j.knosys.2017.07.007
[3]	LI J, YE J, WANG Y, et al. Technical section: fitting 3D garment models onto individual human models[J]. Computers & Graphics, 2010,34(6):742-755.
[4]	WANG L C, ZENG X Y, KOEHL L, et al. Intelligent fashion recommender system: fuzzy logic in personalized garment design[J]. IEEE Transactions on Human-Machine Systems, 2015,45(1):95-109. doi: 10.1109/THMS.2014.2364398
[5]	孙守迁, 徐爱国, 黄琦, 等. 基于角色几何碰撞体估计的实时服装仿真[J]. 软件学报, 2007,18(11):2921-2931.
	SUN S Q, XU A G, HUANG Q, et al. Real-Time garment simulation based on geometrically approximating character with colliding objects[J]. Journal of Software, 2007,18(11):2921-2931.
[6]	孟祥令, 张渭源. 服装压力舒适性的研究进展[J]. 纺织学报, 2006,27(7):109-112.
	MENG Xiangling, ZHANG Weiyuan. Progress of study on pressure comfort of clothing[J]. Journal of Textile Research, 2006,27(7):109-112.
[7]	SONG H K, ASHDOWN S P. Investigation of the validity of 3-D virtual fitting for pants[J]. Clothing and Textiles Research Journal, 2015,33(4):314-330. doi: 10.1177/0887302X15592472
[8]	MENG Y W, MOK P Y, JIN X G. Interactive virtual try-on clothing design systems[J]. Computer-Aided Design, 2010,42(4):310-321. doi: 10.1016/j.cad.2009.12.004
[9]	MENG Y W, MOK P Y, JIN X G. Computer aided clothing pattern design with 3D editing and pattern alteration[J]. Computer-Aided Design, 2012,44(8):721-734. doi: 10.1016/j.cad.2012.03.006
[10]	SAYEM A S M, KENNON R, CLARKE N. 3D CAD systems for the clothing industry[J]. International Journal of Fashion Design, Technology and Education, 2010,3(2):45-53. doi: 10.1080/17543261003689888
[11]	LIU K, KAMALHA E, WANG J, et al. Optimization design of cycling clothes' patterns based on digital clothing pressures[J]. Fibers Polymers, 2016,17(9):1522-1529. doi: 10.1007/s12221-016-6402-2
[12]	LEE W, KO H S. Heuristic misfit reduction: A programmable approach for 3D garment fit customiza-tion[J]. Computers & Graphics-Uk, 2018,71:1-13.
[13]	HINDS B K, MCCARTNEY J J V C. Interactive garment design[J]. Visual Computer, 1990,6(2):53-61. doi: 10.1007/BF01901066
[14]	TAN C, SUN F, KONG T, et al. A survey on deep transfer learning[C]// The 27th International conference on artificial neural networks. Cham: Springer, 2018: 270-279.
[15]	SIMONYAN K, ZISSERMAN A. Very deep convolutional networks for large-scale image recogni-tion[C/OL]. San Diego: Proceedings of the 2015 International Conference on Learning Representations. 2015: 1150-1210. [2020-03-10] https://arxiv.org/abs/1409.1556.htm.
[16]	KRIZHEVSKY A, SUTSKEVER I, HINTON G E. ImageNet classification with deep convolutional neural networks[C]// Proceedings of the 25th international conference on neural information processing systems. New York: Curran Associates, Inc., 2012: 1097-1105.
[17]	CLO Virtual Fashion Inc. CLO 3D[CP/OL]. [2020-03-10]. https://www.clo3d.com/.htm.
[18]	OJALA T, PIETIKAINEN M, MAENPAA T. Multiresolution gray-scale and rotation invariant texture classification with local binary patterns[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2002,24(7):971-987. doi: 10.1109/TPAMI.2002.1017623
[19]	SUYKENS J A K, VANDEWALLE J. Least squares support vector machine classifiers[J]. Neural Processing Letters, 1999,9(3):293-300. doi: 10.1023/A:1018628609742

相关文章 15

[1]	王奕文, 罗戎蕾, 康宇哲. 基于卷积神经网络的汉服关键尺寸自动测量[J]. 纺织学报, 2020, 41(12): 124-129.
[2]	陈咪, 叶勤文, 张皋鹏. 斜裁裙参数化结构模型的构建[J]. 纺织学报, 2020, 41(07): 135-140.
[3]	龚雪, 袁理, 刘军平, 杨亚莉, 刘沐黎, 柯政涛, 鄢煜尘. 混合色彩空间与多核学习的色纺织物组织点识别[J]. 纺织学报, 2020, 41(05): 58-65.
[4]	石康君, 王静安, 高卫东. 基于傅里叶频谱特征的织物平整度客观评级[J]. 纺织学报, 2019, 40(11): 50-56.
[5]	许倩, 陈敏之. 基于深度学习的服装丝缕平衡性评价系统[J]. 纺织学报, 2019, 40(10): 191-195.
[6]	陶开鑫, 俞成丙, 侯颀骜, 吴聪杰, 刘引烽. 基于最小二乘支持向量机的棉针织物活性染料湿蒸染色预测模型[J]. 纺织学报, 2019, 40(07): 169-173.
[7]	朱浩, 丁辉, 尚媛园, 邵珠宏. 多纹理分级融合的织物缺陷检测算法[J]. 纺织学报, 2019, 40(06): 117-124.
[8]	邢文宇, 邓娜, 辛斌杰, 于晨. 基于多特征融合图像分析技术的羊毛与羊绒鉴别[J]. 纺织学报, 2019, 40(03): 146-152.
[9]	李佳平沈国康欧耀明孟想辛斌杰. 应用连续投影算法及最小二乘支持向量机的单组分纺织品识别[J]. 纺织学报, 2018, 39(08): 46-51.
[10]	金关秀张毅楼永平祝成炎. 应用粗糙集和支持向量机的熔喷非织造布过滤性能预测[J]. 纺织学报, 2018, 39(06): 142-148.
[11]	张陆佳林兰天陈春敏申炎仃高琮. 基于主成分分析的纤维拉伸断裂声发射信号识别[J]. 纺织学报, 2018, 39(01): 19-24.
[12]	张建新张银露胡旭东. 光谱优化处理结合多层次支持向量机的混合染液浓度检测方法[J]. 纺织学报, 2017, 38(07): 90-94.
[13]	李东万贤福汪军. 采用傅里叶描述子和支持向量机的服装款式识别方法[J]. 纺织学报, 2017, 38(05): 122-127.
[14]	曹霞李岳阳罗海驰蒋高明丛洪莲. 蕾丝花边的改进型纹理特征检索方法[J]. 纺织学报, 2016, 37(06): 142-154.
[15]	陈孝之谢莉青. 织物颜色配准到标准色卡的计算机识别与仿真[J]. 纺织学报, 2016, 37(05): 150-154.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

模型序列	下半身长	腰围	臀围	膝盖围	小腿长
0	897.7	600.0	834.2	297.8	365.5
1	897.7	619.9	854.3	303.9	356.5
2	938.6	640.0	880.8	313.3	374.9
3	938.5	660.0	896.9	318.8	374.9
4	938.5	681.0	917.8	325.0	374.8
5	979.4	700.0	939.5	333.7	393.3
6	979.4	720.1	958.5	339.6	393.3
7	1 020.1	740.0	980.2	348.1	411.7
8	1 020.2	760.0	1 001.4	354.6	411.7

样板序列	A_CW/%	A_T/%
0	0	12.2
1	0.2	8.0
2	0	7.9
3	0.1	4.5
4	0.2	3.4
5	0	3.2
6	0.1	1.4
7	0.1	1.2
8	0	11.2
…	…	…
68	0	2.1
69	0	2.1
70	0.2	3.9
71	0.1	1.1

评估方法	完备化处理	准确率
评估方法	完备化处理	SVM	NB
HOG	否	0.749	0.674
LBP+HOG	否	0.796	0.779
LBP	否	0.778	0.392
HOG	是	0.811	0.378
LBP+HOG	是	0.768	0.760
LBP(本文方法)	是	0.834	0.467

评估方法	准确率		时间/s
评估方法	SVM	NB	时间/s
CLO 3D^[17]			16
Liu^[2]	0.757	0.705	85
HOG	0.811	0.378	11
LBP+HOG	0.768	0.760	13
LBP(本文方法)	0.834	0.467	12