CPC G10L 25/51 (2013.01) [F21V 23/003 (2013.01); G06F 3/14 (2013.01); G10L 25/90 (2013.01)] | 17 Claims |
3. A computer-implemented audio analysis method comprising:
acquiring a series of feature amounts of an audio signal; and
generating boundary data by inputting the acquired series of feature amounts into a boundary estimation model that has learned relationships between: (i) a series of feature amounts, and (ii) boundary data representative of boundaries, each boundary being between consecutive periods in each of which a chord is continuous, wherein
the boundary data represents, for each of a plurality of time points on a time axis, a likelihood that a time point corresponds to a boundary between consecutive periods.
|