Acoustic music similarity determiner

An apparatus comprising at least one processor and at least one memory including computer program code for one or more programs, the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus to:

• quantise parameters associated with a statistical distribution of audio features of a first audio track;
• quantise parameters associated with a statistical distribution of audio features of at least one further audio track, wherein the parameters associated with the statistical distribution of audio features of the first audio track and the at least one further audio track comprise at least one of:
• mean vector of mel-frequency cepstral coefficients for an audio analysis frame or standard deviation vector of mel-frequency cepstral coefficients for the audio analysis frame;
• select a divergence measure indicating a divergence between the first audio track and the at least one further audio track from a look up table of predetermined divergence values by using the quantised parameters associated with the statistical distribution of audio features of the first audio track and the quantised parameters associated with the statistical distribution of audio features of the at least one further audio track as inputs to the look up table;
• determine a divergence measure threshold value from at least the divergence measure between the statistical distribution of audio features of a first audio track and the statistical distribution of audio features of the at least one further audio track;
• compare the divergence measure with the divergence measure threshold value in order to determine an audio similarity between the first audio track and the at least one further audio track;
• retain the at least one further audio track dependent on whether the divergence measure is less than the divergence measure threshold value;
• and
• identify the retained at least one further audio track is similar to the first audio track.

A computer program product comprising at least one non-transitory computer readable storage medium, the computer readable storage medium comprising a set of instructions, which when executed by one or more processors, causes an apparatus at least to perform:

• quantising parameters associated with a statistical distribution of audio features of a first audio track;
• quantising parameters associated with a statistical distribution of audio features of at least one further audio track, wherein the parameters associated with the statistical distribution of audio features of the first audio track and the at least one further audio track comprise at least one of:
• mean vector of mel-frequency cepstral coefficients for an audio analysis frame or standard deviation vector of mel-frequency cepstral coefficients for the audio analysis frame;
• selecting a divergence measure indicating a divergence between the first audio track and the at least one further audio track from a look up table of predetermined divergence values by using the quantised parameters associated with the statistical distribution of audio features of the first audio track and the quantised parameters associated with the statistical distribution of audio features of the at least one further audio track as inputs to the look up table;
• determining a divergence measure threshold value from at least the divergence measure between the statistical distribution of audio features of a first audio track and the statistical distribution of audio features of the at least one further audio track;
• comparing the divergence measure with the divergence measure threshold value in order to determine an audio similarity between the first audio track and the at least one further audio track;
• retaining the at least one further audio track dependent on whether the divergence measure is less than the divergence measure threshold value;
• and
• identifying the retained at least one further audio track is similar to the first audio track.