In this paper mp algorithm is used to handle the sinusoid modeling. Based on abstract mp algorithm, some improvements and modifications are made with respect to the characteristics of audio signals.
采用MP算法来处理正弦波模型，根据音频信号的特点对抽象的MP算法进行改进和变化。
Psychoacoustics of the human hearing is explicitly included into the algorithm through modifying the MP metric by the time-varying masking threshold of the input signal.
通过计算时变输入信号的心理声学参数，用信号的掩蔽值对匹配跟踪矩阵进行加权。
In addition, the original multi_atoms rapid matching pursuit (MAMP) decomposition algorithm is improved in combination with mp decomposition based on FFT algorithm.
此外，结合FFT的MP分解算法，对原有的多原子匹配追踪(MAMP)分解算法进行改进。
However, the proposed algorithm based on mp can realize high speed, at the same time, the performance of the signal decomposition is assured.
本文研究工作是基于MP分解存在计算量大、过完备原子库存储量大的问题展开的。
However, the proposed algorithm based on mp can realize high speed, at the same time, the performance of the signal decomposition is assured.
本文研究工作是基于MP分解存在计算量大、过完备原子库存储量大的问题展开的。
应用推荐