源代码及论文下载地址：
http://cn.mathworks.com/matlabcentral/profile/authors/5133554-ke-kun-huang
The matlab code written by the authors for the paper: Ke-Kun Huang, Dao-Qing Dai, Chuan-Xian Ren, Zhao-Rong Lai. Learning Kernel Extended Dictionary for Face Recognition. IEEE Transactions on Neural Networks and Learning Systems, 2016, Accepted. http://dx.doi.org/10.1109/TNNLS.2016.2522431
Abstract: Sparse Representation Classifier (SRC) and Kernel Discriminant Analysis (KDA) are two successful methods for face recognition. SRC is good at dealing with occlusion while KDA does well in suppressing intra-class variations. In this paper, we propose Kernel Extended Dictionary (KED) for face recognition, which provides an efficient way for combining KDA and SRC. We first learn several kernel principal components of occlusion variations as an occlusion model, which can represent the possible occlusion variations efficiently. Then the occlusion model is projected by KDA to get the kernel extended dictionary, which can be computed via the same ``kernel trick" as new testing samples. Finally, we use structured SRC for classification, which is fast as only a small number of atoms are appended to the basic dictionary and the feature dimension is low. We also extend KED to multi-kernel space to fuse different types of features at kernel level. Experiments are done on several large-scale datasets, demonstrating that not only does KED get impressive results for non-occluded samples, but it also handles occlusion well without overfitting, even with a single gallery sample per subject.
摘要：稀疏表示分类(SRC)和核判别分析(KDA)是两种人脸识别的好方法。SRC擅长处理遮挡，KDA则能很好的压制类内变化。本文提出核扩展字典(KED)用于人脸识别，提供了结合SRC和KDA的一种有效的途径。首先学习在核空间遮挡变化的前几个主成分作为遮挡模型，使得可以有效地表达可能的遮挡变化。然后用KDA把遮挡模型进行投影以得到核扩展字典，这个过程和一般的核方法一样可以不用显式地使用非线性变换。最后，使用结构化SRC进行分类。因为只增加了少数的原子到基本字典，而且特征维数很低，所以分类很快。我们还把KED扩展到多核空间，使得可以融合多个特征。在几个大规模的人脸数据库中的实验表明，KED不仅能够对无遮挡样本取得很高的识别率，而且能同时很好地处理遮挡而不会过拟合，甚至只用每人一个数据库样本。
评论 5