Method and system for detecting abnormal use of voice input equipment

Abstract

The invention discloses a method and system for detecting abnormal use of voice input equipment. A characteristic extracting mode which is more comprehensive to remote voice signal characteristics and closer to human perception is adopted in the invention so as to roughly judge background voice, normal voice and remote voice. On the basis of classification of audio signals, by adopting the combined method of a modern signal processing technology and a statistical machine learning theory, the problem of multiple limits on front-end voice input by a traditional method is overcome so that signal-level quality scoring is closer to human scoring. According to the invention, the difference problem of front-end input voice quality due to artificial use errors of the equipment in large-scale tests of spoken language is solved.

Description

technical field [0001] The invention relates to the technical field of voice signal processing in the electronic information industry, in particular to a method and system for detecting abnormal use of voice input equipment. Background technique [0002] Voice is an important carrier of information transmission, and its related communication, coding, storage and processing systems have become a necessary means of information exchange in modern society, and have been widely used in various fields of society. Especially in the automatic computer speaking test, the judgment of the quality of the front-end input voice becomes the key to the estimation of the speaking level of the back-end learners. The voice quality of front-end input is usually irrelevant to the level of oral learners. This requires the system to be designed to maintain the same standard and quality as much as possible so that learners can take the exam fairly and impartially. In order to achieve this goal, o...

AI Technical Summary

Problems solved by technology

[0005] 2) The signal-to-noise ratio is too low: During the oral automatic test, due to the physical material and shape of the classroom room, the distance between the microphones violates the requirements, etc., the voices of the surrounding speakers obliterate the target voice, resulting in a low input signal-to-noise ratio;

[0008] Although objective evaluation has the advantages of saving time and effort, it still cannot reflect the ability of the machine to accept the abnormal voice of the device.

There are two bottlenecks restricting the performance of the evaluation module: First, the speech quality evaluation is not enough to distinguish the abnormal speech (especially the far-speaking speech) and the background speech (the MOS value is low), so the system should have the ability to use The abnormal voice of the device, the background voice and the normal voice have a strong ability to distinguish

Images