国际上关于盲源分离技术可以分为两类，瞬时混合信号（Instantaneous Mixtures）的盲分离和卷积混合信号（ConvolutiveMixtures）的盲分离。目前大 多数盲分离问题都是假设观测信号是源信号的即时、线性混合。
基于4个网页-相关网页
并且以线性瞬时混合信号为例，实现了对一般信号、声音信号和图像信号以峰度为判别依据的盲分离和抽取。
And we take the linear instantaneous mixtures as an example, and put it into realty in general signals, sound signals and image signals based on kurtosis.
盲源分离的目的在于只利用观测数据把被瞬时线性混合的源信号恢复出来。
Blind separation of sources consists of recovering a set of signals in which only instantaneous linear mixing are observed.
提出了基于相关性的瞬时线性混合盲信号分离算法。
An algorithm based on the correlation characters for instantaneous linear mixture of the source signal is proposed.
应用推荐