Audio signal classification method and device

Abstract

An audio signal classification method. The method comprises: according to the voice activity of a current audio frame, determining whether to obtain a frequency spectrum fluctuation of the current audio frame and store same in a frequency spectrum fluctuation memory (101); according to whether the audio frame is percussive music or the activity of a historical audio frame, updating the frequency spectrum fluctuation stored in the frequency spectrum fluctuation memory (102); and according to the statistics of some or all valid data of the frequency spectrum fluctuations stored in the frequency spectrum fluctuation memory, classifying the current audio frame as a voice frame or a music frame (103). Further provided is an audio signal classification device.

Description

technical field [0001] The invention relates to the technical field of digital signal processing, in particular to an audio signal classification method and device. Background technique [0002] In order to reduce the resources occupied during video signal storage or transmission, the audio signal is compressed at the sending end and then transmitted to the receiving end, and the receiving end restores the audio signal through decompression processing. [0003] In audio processing applications, audio signal classification is a widely used and important technique. For example, in audio codec applications, the more popular codec is a hybrid codec. Such codecs usually consist of a speech generation model-based coder (such as CELP) and a transform-based coder (such as an MDCT-based coder). At medium and low bit rates, the encoder based on the speech generation model can obtain better speech encoding quality, but the encoding quality for music is relatively poor, while the tran...

AI Technical Summary

Problems solved by technology

[0006] The disadvantage of this speech / music classifier is that on the one hand, the absolute recognition rate of music still needs to be improved; room for improvement

[0008] Although this type of classifier has a high theoretical basis, it usually has high computational or storage complexity, and the implementation cost is high.

Images