Audio copy detection

An apparatus for audio copy detection, comprising processing circuitry configured to:

• extract a first global audio feature of a first audio, and a first audio segment feature sequence of the first audio, the first global audio feature indicating an attribute of the first audio as a whole, and the first audio segment feature sequence indicating time-based local features in the first audio;
• retrieve an approximate global audio feature corresponding to the first global audio feature from an audio retrieval library, the approximate global audio feature being associated with a target audio;
• obtain a second audio segment feature sequence of the target audio;
• determine an audio similarity between the first audio and the target audio based on the first audio segment feature sequence of the first audio and the second audio segment feature sequence of the target audio;
• and
• determine the target audio as an audio copy of the first audio when the audio similarity satisfies a requirement.

A non-transitory computer-readable storage medium storing instructions which when executed by a processor cause the processor to perform:

• extracting a first global audio feature of a first audio, and a first audio segment feature sequence of the first audio, the first global audio feature indicating an attribute of the first audio as a whole, and the first audio segment feature sequence indicating time-based local features in the first audio;
• retrieving an approximate global audio feature corresponding to the first global audio feature from an audio retrieval library, the approximate global audio feature being associated with a target audio;
• obtaining a second audio segment feature sequence of the target audio;
• determining an audio similarity between the first audio and the target audio based on the first audio segment feature sequence of the first audio and the second audio segment feature sequence of the target audio;
• and
• determining the target audio as an audio copy of the first audio when the audio similarity satisfies a requirement.