icalabSignal
所属分类：matlab编程
开发工具：matlab
文件大小：11988KB
下载次数：24
上传日期：2020-06-14 01:47:31
上 传 者：
mobys
说明： 预处理工具包括：主成分Analysi 小号（PCA），白化，过滤：高通滤波（HPF），低通滤波（LPF），子带滤波器（巴特沃斯，切比雪夫，椭圆）用的滤波器，频率子带和可调顺序子带的数量）或用户定义的 预处理功能。
后处理工具实际上包括：通过去除不想要的组件，噪声或伪像，对原始原始数据进行压缩和重建（“清理”）。
该算法不仅可以执行ICA，还可以执行二阶统计盲源分离（BSS），稀疏分量分析（SCA），非负矩阵分解（NMF），平滑分量分析（SmoCA），因子分析（FA）和任何其他可能的矩阵X = HS + N或 Y = WX形式的因式分解，其中H = W +是混合矩阵或基本向量矩阵。X是观测数据的矩阵，S是原始数据的矩阵，N表示其他噪声的矩阵。
ICA / BSS算法是纯数学公式，功能强大，但机械程序却很复杂：机械得到最佳实施后，用户要做的工作不多。ICALAB的成功和有效使用在很大程度上取决于先验知识，常识以及对预处理和后处理工具的适当使用。
(icalab toolboxes for signal process, based on the software MATLAB.)
文件列表:
ICALABSPv2_2\acrsobibpf.p (10381, 2004-02-24)
ICALABSPv2_2\acsobiro.m (6091, 2004-02-23)
ICALABSPv2_2\acsobiro.p (7287, 2004-02-24)
ICALABSPv2_2\adv_param.p (28812, 2004-02-24)
ICALABSPv2_2\amuse.m (1477, 2004-02-23)
ICALABSPv2_2\ard2ica.p (24786, 2004-02-24)
ICALABSPv2_2\ardkica.p (22765, 2004-02-24)
ICALABSPv2_2\arngica.p (21778, 2004-02-24)
ICALABSPv2_2\benchmarks (0, 2004-02-23)
ICALABSPv2_2\benchmarks\10halo.mat (480192, 2002-08-15)
ICALABSPv2_2\benchmarks\25speakersHALO.mat (1200184, 2002-02-18)
ICALABSPv2_2\benchmarks\25speakersNOSYNC.mat (1200184, 2002-02-18)
ICALABSPv2_2\benchmarks\64sounds_std.mat (1792184, 2002-02-18)
ICALABSPv2_2\benchmarks\ABio7.mat (280200, 2002-08-15)
ICALABSPv2_2\benchmarks\AC10-7sparse.mat (80208, 2002-08-15)
ICALABSPv2_2\benchmarks\ACPos24sparse10.mat (160240, 2002-11-01)
ICALABSPv2_2\benchmarks\ACposin10.mat (160248, 2002-11-01)
ICALABSPv2_2\benchmarks\ACposvsparse.mat (80184, 2002-11-01)
ICALABSPv2_2\benchmarks\ACpwcon10.mat (80184, 2003-03-19)
ICALABSPv2_2\benchmarks\ACsin10d.mat (80184, 2002-08-15)
ICALABSPv2_2\benchmarks\ACsin4d.mat (32216, 2002-08-15)
ICALABSPv2_2\benchmarks\ACsincpos10.mat (80184, 2002-11-01)
ICALABSPv2_2\benchmarks\ACsparse10.mat (80184, 2002-08-15)
ICALABSPv2_2\benchmarks\acspeech16.mat (448184, 2002-08-15)
ICALABSPv2_2\benchmarks\acv10_sin.mat (80184, 2003-02-14)
ICALABSPv2_2\benchmarks\ACvsparse10.mat (80184, 2002-08-15)
ICALABSPv2_2\benchmarks\bbb4.mat (32184, 2003-02-14)
ICALABSPv2_2\benchmarks\ccc4.mat (32184, 2003-02-14)
ICALABSPv2_2\benchmarks\dur3.mat (2646184, 2003-02-14)
ICALABSPv2_2\benchmarks\EEG19.mat (78016, 2002-08-21)
ICALABSPv2_2\benchmarks\Gnband.mat (80192, 2002-08-15)
ICALABSPv2_2\benchmarks\nband5.mat (80184, 2002-08-15)
ICALABSPv2_2\benchmarks\octave3.mat (2646184, 2003-02-14)
ICALABSPv2_2\benchmarks\Sergio7.mat (560192, 2002-08-15)
ICALABSPv2_2\benchmarks\sin10.mat (160184, 2003-10-16)
ICALABSPv2_2\benchmarks\sp4.mat (160184, 2003-02-14)
ICALABSPv2_2\benchmarks\SparSergio7.mat (560184, 2002-11-03)
ICALABSPv2_2\benchmarks\Speech10.mat (640192, 2002-08-15)
ICALABSPv2_2\benchmarks\Speech20.mat (560184, 2002-08-15)
... ...
ICALAB for Signal Processing - BENCHMARKS August 15, 2002
The directory Benchmarks contains typical signals and images
for testing and comparison of various algorithms for
ICA, BSE and BSS.
The most interesting benchmarks are briefly described below.
1.
ACsin10d.mat - contains 10 sine-wave sources of the following form:
s_n = sin ((2n-1)*w*k) for n = 1,2,....,10.
The sources can be easily separated by the second orders
statistics methods (SOS) like AMUSE, EVD or SOBI algorithms.
However, the higher order statistics (HOS)
ICA algorithms fail to reconstruct such sources, because
they are dependent.
It is interesting to note that different ICA algorithms
(for example JADE and Natural gradient algorithms) give usually
different (inconsistent) results ("independent components")
for this benchmark.
Acsin4d.mat benchmark is similar to acsin10d.mat
but it contains only 4 sources.
2.
ACsparse10.mat - contains 10 sparse (smooth bell-shape) sources
that are approximately independent.
The SOS blind source separation algorithms fail
to separate such sources.
Also some ICA algorithms have failed to separate such sources.
Please try ICALAB to compare performance of various algorithms.
3.
ACvsparse10.mat - contains 10 very sparse (short regular pulses).
Second order BSS algorithms fail to separate these sources.
4.
ABio7.mat - this benchmark contains 7 typical biological sources.
This benchmark was proposed by Allan Barros.
5.
Sergio7.mat - this benchmark contains 7 sources (some of them
are asymmetric distributed).
This benchmark was proposed by Sergio Cruces.
6.
AC10-7sparse -contains 10 sensors signals which are mixture
of 7 sources (extracted from the file ACsparse10.mat).
7.
acspeech16.mat contains 16 typical speech signals which have
temporal structure but they are not precisely independent.
Similar benchmarks are :
Speech4.mat, Speech8.mat. Speech10.mat and Speech20.mat with
4, 8, 10 and 20 sounds (speech and music) sources.
10halo.mat contains 10 speech signals highly correlated
(the all 10 speakers say the same sentence).
8.
nband5.mat contains 5 narrow band sources.
This is standard "easy" benchamark.
9.
Gnband.mat contains 5 fourth order colored sources with distribution close
to Gaussian. This is rather "difficult" benchmark. Please try program JADE-TD
to separate the signals from their mixture.
10.
EEG19.mat consists 19 EEG signals with clear heart, eye movements
and eye blinking artifacts.
近期下载者：
相关文件：
收藏者：