Audio detect method, device, apparatus, and computer-readable storage medium

Abstract

The invention discloses an audio detection method, an apparatus, a device and a computer-readable storage medium, wherein, the method comprises the following steps of: obtaining a training sample setcomprising a preset amount of training samples and sample tags of each of the training samples, wherein the training sample comprises audio samples corresponding to the sample tags of the spliced audio and the original audio respectively; training a pre-created convolution neural network by using the training sample set to obtain an audio recognition model; the audio sample with unknown sample label is obtained as the sample to be tested, and the sample to be tested is input to the audio recognition model, so as to obtain the recognition result that the sample to be tested output by the audiorecognition model is mosaic audio or original audio. Audio detection base on a convolution neural network is realized in that application, and the audio detection method is not only prove to have highaccuracy by experiments, but also has no special requirements for training samples and sample to be tested, and is good in generalization.

Description

technical field [0001] The present invention relates to the technical field of audio detection, and more specifically, to an audio detection method, device, equipment and computer-readable storage medium. Background technique [0002] In order to ensure audio security, it is usually necessary to detect it to determine whether the audio is a tampered spliced ​​audio or an untampered original audio. The technical solution for audio detection in the prior art is usually to perform correlation detection on the features of adjacent audio segments, and then determine whether to splice the audio or the original audio, but this method has high requirements on the audio, and requires the audio to have a long time and noise. Small, etc., so it cannot be widely applied to audio under different conditions. [0003] To sum up, the technical solutions for implementing audio detection in the prior art have the problem of poor generalization. Contents of the invention [0004] The purpo...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to provide an audio detection method, device, equipment and computer-readable storage medium, which can solve the problem of poor generalization existing in the technical solutions for audio detection in the prior art

Images