Audio frequency copy detection method based on similarity

Abstract

The invention belongs to the technical field of audio frequency information processing and in particular relates to an audio frequency copy detection method based on similarity. The method comprises the following steps: firstly establishing a reference audio frequency signal database, segmenting the signals before entering the database so that the signals have equal length, wherein the reference audio frequency signals in the database are illegal or bad information; copying and detecting the audio frequency signal according to the established reference audio frequency database, namely, orderly extracting features, generating a histogram and computing the similarity; then judging whether the input audio signal has a copy containing the reference audio frequency in the database by a parallel algorithm so as to obtain the output result, namely the result of whether the audio frequency has illegal or bad information. The method provided by the invention can be used for detecting and filtering unhealthy, violent and retroactive audio on the internet and various audio frequency copy detection application systems to prevent various bad contents from spreading.

Description

technical field [0001] The invention belongs to the technical field of audio information processing, and in particular relates to an audio copy detection method. Background technique [0002] The advancement of audio compression technology and the emergence of large-capacity storage have led to the emergence of massive audio information on the Internet. These audio information are widely used in education, entertainment, news, advertising and other fields, and become an important part of people's daily life. However, a series of problems followed: because the Internet audio is very flexible and open in the production, distribution, playback, transmission and other links, the number of illegal audio and pirated audio content on the Internet is increasing, which seriously damages content providers and related parties. The interests of other parties have hindered the healthy and orderly development of the network audio industry. At the same time, Internet audio has also become...

AI Technical Summary

Problems solved by technology

However, a series of problems followed: because the Internet audio is very flexible and open in the production, distribution, playback, transmission and other links, the number of illegal audio and pirated audio content on the Internet is increasing, which seriously damages content providers and related parties. party interests, hindering the healthy and orderly development of the network audio industry

However, using this technique, the search becomes impractical due to the computational complexity of long-term (say, up to several days) suspect audio signals or many reference audio signals

Of course, one can improve the speed by rough matching, but this will inevitably reduce the correctness and stability of the search

Images