《计算机应用研究》|Application Research of Computers

改进随机子空间LDA结合多补丁集成学习的鲁棒人脸识别算法

Robust face recognition algorithm based on multiple patch integration learning and improved random subspace LDA

免费全文下载 (已被下载 次)  
获取PDF全文
作者 冯代高,张友俊
机构 上海海事大学 信息工程学院,上海 201306
统计 摘要被查看 次,已被下载
文章编号 1001-3695(2019)08-067-2556-05
DOI 10.19734/j.issn.1001-3695.2018.03.0190
摘要 针对化妆对人脸识别准确率的负面影响,提出了基于补丁集成学习的改进鲁棒人脸识别算法。将每张人脸图像嵌入补丁中,并用一组特征描述符描述每个补丁,即本地梯度Gabor模式(LGP)、Gabor空间定序定比测量直方图(HGSFRM)和密集采样局部多值模式(DSLMP)。使用改进的随机子空间线性判别分析(SRS-LDA)方法采样补丁,并在化妆之前与化妆之后图像之间建立多个公共子空间进行集成学习。利用协作和稀疏表示分类器比较这个子空间中的特征向量,同时通过求和规则联合得到分数。实验将提出的算法在多种化妆数据集上进行评估分析,结果表明提出的算法相比于其他专为妆后人脸识别设计的算法有更高的识别精度。
关键词 人脸识别; 鲁棒评估; 描述符算法; 补丁集成; 改进随机子空间
基金项目 国家自然科学基金资助项目(61475137)
本文URL http://www.arocmag.com/article/01-2019-08-067.html
英文标题 Robust face recognition algorithm based on multiple patch integration learning and improved random subspace LDA
作者英文名 Feng Daigao, Zhang Youjun
机构英文名 College of Information Engineering,Shanghai Maritime University,Shanghai 201306,China
英文摘要 Aiming at the negative effects of makeup on face recognition, this paper presented an improved robust face recognition algorithm based on patch integration learning. In the proposed scheme, this paper tessellated each face image into patches and represented each patch with a set of feature descriptors, viz, local Gabor pattern(LGP), histogram of Gabor space fixed ratio measures(HGSFRM) and densely sampled local multi-valued pattern(DSLMP). Then, it used an improved random subspace linear discriminant analysis(SRS-LDA) method to perform ensemble learning by sampling patches and constructing multiple common subspaces between before-makeup and after-makeup facial images. Finally, it used collaborative-based and sparse-based representation classifiers to compare feature vectors in this subspace and the resulting scores combined by the sum-rule. It evaluated the proposed face matching algorithm on multiple makeup dataset. Results show that this algorithm has higher recognition accuracy than other algorithms designed for face recognition under makeup situation.
英文关键词 face recognition; robust evaluation; signature integration algorithm; patch integration; improved random subspace
参考文献 查看稿件参考文献
 
收稿日期 2018/3/22
修回日期 2018/5/7
页码 2556-2560
中图分类号 TP391.41;TP301.6
文献标志码 A