Title：Voice activity detection method and system applied to low-power-consumption circuit

Country：China

Patent No.：202110755667.2

Legal Status：Under review

Inventor：Xian Tang, Bowen Sun

Assignee：International Graduate School at Shenzhen, Tsinghua University

Address：Tsinghua Campus, Xili University Town, Nanshan District, Shenzhen City 518055, Guangdong Province

Filing Date：2021-07-05

Issue Date：

The invention discloses a voice activity detection method and system applied to a low-power-consumption circuit. The method comprises the steps of receiving input voice, carrying out voice feature extraction, adopting sub-band energy features, and reducing the number of sub-bands; obtaining feature values through feature extraction, and then entering subsequent classification; and selecting a linear classifier of a support vector machine for training classification, and outputting a voice classification result to complete voice activity detection. The invention provides a relatively simple voice activity detection method, and relatively high classification accuracy under a low signal-to-noise ratio is obtained by adopting relatively low implementation complexity; and meanwhile, the voice feature extraction part is realized by adopting a full-analog circuit, the implementation possibility of an actual circuit is considered during design, and the requirement of a low-power-consumption circuit can be additionally met.

Patent Certificate： PDF/Jpg