Active voice detection method and system based on noise scene recognition

Abstract

The invention discloses an active voice detection method based on noise scene recognition, and the method comprises the steps: extracting an optimal feature facing a noise classification task from anaudio signal, inputting a feature value into a noise type classifier, so as to recognize a noise type in the audio signal, according to the noise type, determining a preferred feature and a classifiersuitable for the voice and noise classification task, extracting optimal features oriented to the voice and noise classification task from the audio signals, inputting the optimal features into a voice and noise classifier, and judging whether the voice signals exist in the audio signals or not. According to activity voice detection system based on noise scene recognition, before binary classification of noisy voice and noise signals is carried out, according to the method, the current noise type is detected and identified, the most distinctive feature combination is preferably selected for the specific noise type, model parameters can be designed for the specific noise type, and the effectiveness and stability of the performance of the whole detection process under different noise typesare ensured.

Description

technical field [0001] The invention relates to the technical field of voice data processing, in particular to a noise scene recognition-based active voice detection method and system. Background technique [0002] There are often pauses and intermittent phenomena in a segment of speech signal. These "silent" segments will be superimposed with environmental noise and do not contain effective speech information. This type of information occupies a large data transmission resource and interferes with the effect of speech signal processing. The goal of active voice detection (voice activity detection, VAD) technology is to detect real voice passages from the signal and remove these "silent" parts, thereby reducing the burden of subsequent voice signal processing. Therefore, active voice detection technology is widely used It is used in speech coding, speaker recognition, automatic speech recognition, abnormal sound detection and other systems. [0003] In view of the wide appl...

AI Technical Summary

Problems solved by technology

[0004] Supervised methods are mainly composed of feature extraction and classifier design. In terms of feature extraction, in order to effectively distinguish the acoustic characteristics of noise and speech signals, researchers currently extract high-dimensional features from different angles, such as energy features, process Zero rate features, Mel Frequency Cepstrum Coefficient (MFCC) features, fuzzy entropy features, autocorrelation coefficient features, wavelet coefficient features, etc., and combine multiple features to fuse multi-angle feature information, however , although these feature combinations have a certain ability to distinguish noise and speech under specific noise types, in practical applications, due to the time-varying nature of noise types, it is often difficult for feature combinations set for general situations to show stability in dynamic noise scenarios. The ability to distinguish, and the high dimensionality of features often brings a burden to the subsequent use of classifiers

Images