A method and device for extracting multiple fundamental frequencies

Abstract

The invention discloses a multi-fundamental frequency extraction method and a multi-fundamental frequency extraction device based on empirical mode decomposition and a hidden Markov model. The method comprises steps: an auditory filter bank is used for filtering a speech signal, and framing is carried out on the signal after filtering; an auto-correlation function is calculated on each time frequency unit for an auditory spectrum; on the basis of an intrinsic mode function obtained through the empirical mode decomposition, the instantaneous frequency of each time frequency unit dominant sound source is calculated; on the basis of each instantaneous frequency, a frequency matching function is calculated; the frequency matching function is used for building the likelihood probability of each fundamental frequency state, and a corpus is used for counting the transition probability between each fundamental frequency state and a fundamental frequency value; and the likelihood probability of each fundamental frequency state is enhanced, the enhanced likelihood probability is combined with the corresponding transition probability, and the hidden Markov model is used for extracting a multi-fundamental frequency track of the speech signal.

Description

technical field [0001] The invention relates to the decomposition of the empirical mode of digital signal processing, the analysis of the speech signal filter group, the extraction of the pitch frequency of the speech signal, the construction of the likelihood probability and transition probability of the hidden Markov model. Background technique [0002] Pitch extraction and track tracking are of great significance to many speech and audio signal processing technologies, such as audio retrieval and classification, Chinese intonation recognition, and single-channel speech separation technology. Some well-performed pitch extraction algorithms exist for detecting a single pitch in clean or slightly noisy speech. However, the assumption of a single fundamental frequency makes this type of algorithm unable to be used in the case where there are multiple fundamental frequencies in speech at the same time, such as the situation where two speakers speak at the same time or the situ...

AI Technical Summary

Problems solved by technology

However, in the case of multi-pitch extraction, there may be high-order harmonics with similar energy but belonging to different pitch frequencies in a time-frequency unit, which will cause its amplitude modulation rate to not belong to any pitch frequency. In the case of harmonics, this will cause the peak height and peak position of the corresponding autocorrelation function to be erroneously shifted, thereby negatively affecting the extraction of multi-pitch frequencies

Images