Special audio event layered and generalized identification method based on SVM (Support Vector Machine) and GMM (Gaussian Mixture Model)

Abstract

The invention relates to a special audio event layered and generalized identification method based on a combination of an SVM (Support Vector Machine) and a GMM (Gaussian Mixture Model), and belongs to the technical field of a computer and audio event identification. The special audio event layered and generalized identification method comprises the following steps of: firstly, obtaining an audio characteristic vector file of a training sample; secondly, respectively carrying out model training on a great quantity of audio characteristic vector files (of the training samples) with various types by using a GMM method and an SVM method, so as to obtain the GMM model with generalization capability and an SVM classifier, and complete offline training; and finally, carrying out layered identification on the audio characteristic vector files to be identified by using the GMM model and the SVM classifier. With the adoption of the method provided by the invention, the problems that the conventional special audio event identification is low in identification efficiency on a continuous audio stream, very short in continuing time, high in audio event false dismissal probability can be solved. The method can be applied to searching a special audio and monitoring a network audio based on contents.

Description

technical field [0001] The invention relates to a layered and generalized recognition method for a specific audio event that combines a support vector machine (SVM) and a Gaussian mixture model (GMM), and belongs to the technical field of computer and audio event recognition. Background technique [0002] A specific audio event is an audio segment designated by the user with a specific semantic or content. [0003] With the rapid development of computer and network technology, the rapid expansion of audio and video files and streaming media data, it is becoming more and more important to quickly and accurately find or identify the required specific audio events from the massive audio information. Due to the urgent needs of applications such as traffic monitoring and security monitoring in sensitive areas, specific audio event recognition technology has been extensively studied in recent years. [0004] Specific audio event recognition technology is to identify specific audi...

AI Technical Summary

Problems solved by technology

[0019] The purpose of the present invention is to solve the problems of low audio event recognition accuracy, short-duration audio event segment missing detection rate, and slow audio event recognition speed in continuous audio event streams, and propose a specific audio event that combines SVM and GMM. Hierarchical generalization of event recognition method, through the use of audio feature combinations such as MFCC, and the fusion of SVM classifier and GMM model to achieve high-precision and rapid recognition of specific audio event hierarchies and generalizations

Images