Aiming at the high alarm rate and poor robustness in the existing intrusion detection algorithms, this paper proposes a cooperative intrusion detection algorithm based on sparse representation.
针对现有入侵检测算法误报率较高和鲁棒性较差的问题，提出一种基于稀疏表示的协同入侵检测算法。
参考来源 - 基于稀疏表示的协同入侵检测算法The graph in SDA is constructed by sparse representation and incorporate semi-supervised discriminant analysis(SSDA),thus the local structure information is automatically modeled,and with the natural discriminative power of sparse representation,SDA can get better performance only resorting to few extra unlabeled samples.
SDA引入稀疏表述并结合半监督判决分析SSDA,通过稀疏重构处理,获得图像的局部结构信息,且由于稀疏表述本身具有的判决性,SDA只需很少样本就能获得较好的效果。
参考来源 - 稀疏判决分析在表情识别中的应用In order to maximize the robustness to occlusion, we introduce transverse partition scheme, apply random projection and sparse representation to each of the blocks and aggregate the results by mean value of the discrimination factor.
同时,为了达到对于遮掩的极大鲁棒性,采取图像横向四等分策略,在各等分块上分别进行随机投影和稀疏表征,以四块平均判据值进行分类识别。
参考来源 - 基于随机投影和稀疏表征的红外人脸识别方法Experimental results show the result is superior to that of wavelet’s. For singularities alone curves, curvelets can give sparse representation.
对具有曲线奇异性的目标，曲线波能对其进行稀疏表示。
参考来源 - 多尺度分析在图像处理中的应用研究Sparse component analysis (SCA) is a method of signal processing based on sparse representation.
一般的信号在时域中并不是稀疏的，因此在很多应用中需要寻找有效的稀疏表征来进行信号处理。
参考来源 - 欠定混叠语音信号盲分离方法研究·2,447,543篇论文数据，部分数据来源于NoteExpress
This paper discusses the recoverability of underdetermined blind source separation(BSS), based on a two-stage sparse representation approach.
基于一种两步稀疏表示的方法，利用随机框架讨论欠定盲源分离的恢复能力。
The modified algorithm reduces the influence of noise on L-R algorithm effectively by using de-noise algorithm based on adaptive sparse representation.
这种改进算法通过使用基于自适应稀疏表示的去噪算法有效地减少了噪声对L-R算法的影响。
Considering the good performance of sparse representation for image classification, a novel sparse representation based text detection method is presented.
针对稀疏表示具有良好的分类特性，本文提出了一种基于稀疏表示的文字检测方法。
应用推荐