一种基于产生式分数空间的单样本人脸识别方法

摘要/Abstract

摘要：

提出一种基于产生式分数空间的单样本人脸识别方法.首先设计了适用于人脸表示的概率产生式模型，并有效地结合了分部式方法的灵活性和稀疏成分分析的稳健性.然后基于该模型导出分数函数（特征映射），并构建了本质上是观测数据、隐变量和模型参数函数的概率相似度.最后，基于2个标准人脸数据库进行了仿真实验，验证了所提出方法的有效性.

关键词:

产生式分数空间；单样本人脸识别；概率相似度

Abstract:

In this paper, a generativescorespace model was proposed, based on which a face recognition approach was derived. First, the proposed approach designed a probabilistic generative model for face representation, which effectively combined the flexibility of partsbased paradigm with the robustness of sparse component analysis. Then, a score function (i.e. feature mapping) was derived based on the model. Besides, a similarity measure was constructed for single sample face identification, which is essentially the function over observed variables, hidden variables and model parameters. The proposed approach was evaluated on two standard face databases to validate its effectiveness.

Key words:

generative score space; single sample face identification; probabilistic similarity measure

中图分类号:

TP 391.4

王斌1，刘允才2，茅红伟1. 一种基于产生式分数空间的单样本人脸识别方法[J]. 上海交通大学学报, 2017, 51(2): 202-.

WANG Bin1,LIU Yuncai2,MAO Hongwei1. Single Sample Face Identification Based on Generative Score Space[J]. Journal of Shanghai Jiao Tong University, 2017, 51(2): 202-.

参考文献

［1］王科俊，段胜利，冯伟兴. 单训练样本人脸识别技术综述［J］. 模式识别与人工智能，2008，21(5)：635-642.
WANG Kejun, DUAN Shengli, FENG Weixing. A survey of face recognition using single training sample ［J］. Pattern Recognition and Artificial Intelligence, 2008, 21(5): 635-642.
［2］PERINA A, CRISTANI M, CASTELLANI U, et al. Free energy score spaces: Using generative information in discriminative classifiers ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012, 34(7): 1249-1262.
［3］LI X, LEE T, LIU Y. Hybrid generativediscriminative classification using posterior divergence ［C］//IEEE Conference on Computer Vision and Pattern Recognition. Colorado Springs, USA: IEEE, 2011: 2713-2720.
［4］WANG H, ZHENG W. Robust sparsitypreserved learning with application to image visualization ［J］. Knowledge and Information Systems, 2014, 39(2): 287-304.
［5］HINTON G H. Training products of experts by minimizing contrastive divergence ［J］. Neural Computation, 2002, 14(8): 1771-1800.
［6］GREG C, WEI G, TANNER M A. A Monte Carlo implementation of the EM algorithm and the poor man’s data augmentation algorithms ［J］. Journal of the American Statistical Association, 1990, 85(411): 699-704.
［7］GILKS W R, WILD P. Adaptive rejection sampling for Gibbs sampling ［J］. Applied Statistics, 1992, 41(2): 337-348.
［8］ZHANG G, WANG Y. Robust 3D face recognition based on resolution invariant features ［J］. Pattern Recognition Letters, 2011, 32(7): 1009-1019.
［9］SHANG F, JIAO L C, WANG F. Graph dual regularization nonnegative matrix factorization for coclustering ［J］. Pattern Recognition, 2012, 45(6): 2237-2250.
［10］SIM T, BAKER S, BSAT M. The CMU pose, illumination, and expression (PIE) database［C］//IEEE International Conference on Automatic Face and Gesture Recognition. Seoul, Korea: IEEE, 2002, 25(12): 46-51.
［11］DU B, ZHANG L. A discriminative metric learning based anomaly detection method ［J］. IEEE Transactions on Geoscience and Remote Sensing, 2014, 52(11): 6844-6857.
［12］CAI D, HE X, HAN J, et al. Graph regularized nonnegative matrix factorization for data representation ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2010, 33(8): 1548-1560.
［13］ROWEIS S, SAUL L. Nonlinear dimensionality reduction by locally linear embedding ［J］. Science, 2000, 290(5500): 2323-2326.
［14］BELHUMEUR P N, HESPANHA J P, KRIGMAN D J. Eigenfaces vs. fisherfaces: Recognition using class specific linear projection ［J］.IEEE Transactions on Pattern Analysis and Machine Intelligence, 1997，19(7)： 711-720.
［15］ZHANG Y, YRUNG D Y. Semisupervised generalized discriminant analysis ［J］. IEEE Transactions on Neural Networks, 2011, 22(8): 1207-1217.

[1]	白雄飞, 龚水成, 李雪松, 许博, 杨晓力, 王明彦. 基于泊松融合数据增强的焊缝金相组织缺陷分类研究[J]. 上海交通大学学报, 2023, 57(10): 1316-1328.
[2]	高涛, 文渊博, 陈婷, 张静. 基于窗口自注意力网络的单图像去雨算法[J]. 上海交通大学学报, 2023, 57(5): 613-623.
[3]	叶仙，胡洁，田畔，戚进，车大钿，丁颖. 基于精细复合多尺度熵与支持向量机的睡眠分期[J]. 上海交通大学学报（自然版）, 2019, 53(3): 321-326.
[4]	刘宇，江宏毅，王仕亮，王伊冰，陈燕苹. 基于加速度时域特征的实时人体行为模式识别[J]. 上海交通大学学报（自然版）, 2015, 49(02): 169-172.
[5]	王佳婧1，2，张树生1，何卫平1. 基于多信息融合的失效DPM码识别[J]. 上海交通大学学报（自然版）, 2014, 48(12): 1675-1680.
[6]	赵洪伟，谢永芳，曹斌芳，蒋朝辉. 基于Gabor小波和LPP的浮选过程泡沫纹理特征提取及应用[J]. 上海交通大学学报（自然版）, 2014, 48(07): 942-947.
[7]	贾松敏1,王爽1,王丽佳1,2,李秀智1. CTF定位策略下基于多特征的智能机器人目标跟踪[J]. 上海交通大学学报（自然版）, 2014, 48(07): 1039-1045.
[8]	王佳婧1,2，张树生1，何卫平1. 金属零件上二维条码高光区域的恢复[J]. 上海交通大学学报（自然版）, 2013, 47(12): 1836-1841.
[9]	方挺，董冲，胡兴柳，王彦. 航拍图像中绝缘子串的轮廓提取和故障检测[J]. 上海交通大学学报（自然版）, 2013, 47(12): 1818-1822.
[10]	刘洪涛，梁振宁，胡文，莫锦秋，王石刚. 基于局部特征点检测与匹配的微悬臂梁变形受力测量方法[J]. 上海交通大学学报（自然版）, 2013, 47(12): 1842-1847.
[11]	王晓端1，刘书炘2，刘满华1. 基于模型分析的指纹奇异点检测[J]. 上海交通大学学报（自然版）, 2013, 47(08): 1222-1226.
[12]	胡文1, 刘洪涛1, 胡春2, 范明霞3, 莫锦秋1, 王石刚1. 基于伪Zernike矩不变量分析的视觉测力[J]. 上海交通大学学报（自然版）, 2013, 47(04): 589-593.
[13]	刘洪涛, 蒋如意, 胡文, 莫锦秋, 王石刚. 基于路面垂直轮廓B样条拟合的车辆运动估计偏移抑制[J]. 上海交通大学学报（自然版）, 2013, 47(04): 630-634.
[14]	李斌, 叶昊. 基于仿射几何不变量的鲁棒图像配准算法[J]. 上海交通大学学报（自然版）, 2012, 46(12): 1881-1884.
[15]	吕复强1, 王慧1, 刘泓2. 基于改进混合高斯模型的交通流背景检测算法[J]. 上海交通大学学报（自然版）, 2012, 46(12): 1920-1925.