Cooperative FFT Modeling of Sparse Analysis of Traditional Opera Audio Signals and Piano Impromptu Accompaniment

This paper studies the cooperative FFT modeling of traditional opera audio signal sparse analysis and piano impromptu accompaniment. By analyzing the structural characteristics of the over-complete atomic library used in traditional opera speech signal sparse decomposition, an improved signal sparse decomposition algorithm is proposed. According to the characteristics of the voice signal, the algorithm reduces the audio signal of piano improvisation based on the sparse decomposition realized by the FFT algorithm, thus not only further improving the decomposition speed, but also expressing the voice signal in a more sparse form, combined with the FFT of piano improvisation accompaniment Modeling improves effectiveness by 7.3%.