南京大学学报(自然科学版) ›› 2012, Vol. 48 ›› Issue (1): 7076.
陈才扣**,喻以明,史俊
Chen Cai一Kou ,Yu Yi -Ming,Shi Jun
摘要: 基于稀疏表示的分类器(sparse representatiorrbased classifier, SRC)被证实是一种非常有效的分类器.但SRC往往要通过一个超完备基来求得测试样本的稀疏表示,当数据库的数据量较大时,算
法的计算复杂度成为限制其优良性能的瓶颈,致使SRC无法用于实时识别.钊一对该问题,提出一种简便有效的改进算法,其试图寻求一个较小的超完备基来计算测试样本的稀疏表示,从而大大的缩减算法的
计算复杂度.其体来说,对于每个测试样本点,首先,求出该测试样本点可能归属的类别,而后利用可能归属类的样本而并非所有的训练样本来对测试样本进行稀疏表示计算.()RI.人脸库和FFRF’I}人脸库
上的实验结果表明改进算法不仅能较大程度的缩减算法的计算复杂度,而且排除了十扰类的影响,从而在某种程度上提高了算法的识别率.
[1]Donoho D. Compressive sensing. IEEE Trans actions on Information Theory, 2006,52(4):1289~1306 [2]Yang A Y,Wright J,Ma Y,et al. Feature se- lection in face recognition; A sparse representa tion perspective.Technique Report, 2007. [3]Candes E J,Wakin M Y. An introduction to compressive sensing, IEEE Signal Processing Magazine,2008,24(4),118一120,124. [4]Cheng B, Yang J C,Yan S C, et al. Learning with l1-graph for image analysis, IEEE Trans- actions on Image Processing, 2010,19(4): 858一866. [5]Yang J C, Wright J,Huang T,et al. Image super-resolution as sparse representation of raw image patches. IEEE Conference on Computer Vision and Pattern Recognition, 2008,1一8. [6]Elad M, Aharon M, image denoising via sparse and redundant representations over learned dic tionaries. IEEE Transactions on Image Process- ing, 2006,15(6):3736一3745. [7]Wright J,Yang A,Ganesh A,et al. Robust face recognition via sparse representation, IEEE Transactions on Pattern Analysis and Machine Intelligence, 2009,31(2):210一227 [8]Li C G, Guo J,Zhang H G. Local sparse repre sentation based classification.The 20th Intcrna tional Conference on Pattern Recognition IEEE,2010,649一652 [9]Aharon M, Elad M,Bruckstcin A. K一SVD: An algorithm for designing overcomplete dir tionarics for sparse representation, IEEE Transactions on Signal Processing,2006,54 (11):4311一4322. [10]Huang J B, Yang M H. Fast sparse representa tion with prototypes, IEEE conference on Com puter Vision and Pattern Recognition, 2010 3218一3625. [11]Gheng H,Liu Z C,Yang J. Sparsity lnduced Similarity Measure for Label Propagation, IEEE conference on Computer Vision, IEEE, 2009,317一324. [12]Sparselab matlab software package,http// sparsclab. Stanford. edu/. |
No related articles found! |
|