Audio segmentation and classification
AssigneeMicrosoft Corporation
InventorsHao Jiang, Hongjiang Zhang
Life of this patent
Estimating the term…
Abstract
A portion of an audio signal is separated into multiple frames from which one or more different features are extracted. These different features are used, in combination with a set of rules, to classify the portion of the audio signal into one of multiple different classifications (for example, speech, non-speech, music, environment sound, silence, etc.). In one embodiment, these different features include one or more of line spectrum pairs (LSPs), a noise frame ratio, periodicity of particular bands, spectrum flux features, and energy distribution in one or more of the bands. The line spectrum pairs are also optionally used to segment the audio signal, identifying audio classification changes as well as speaker changes when the audio signal is speech.
Claims9
A method comprising:
- receiving an audio signal;
- separating the audio signal into a plurality of portions;
- classifying each of the plurality of portions, based at least in part on periodicity features of the portion, as one of:
- speech, music, silence, and environment sound;
- extracting line spectrum pairs from each of the plurality of frames;
- generating an input Gaussian Model corresponding to the plurality of frames based on the extracted line spectrum pairs;
- identifying one of the plurality of trained Gaussian Models that is closest to the input Gausssian Model;
- determining a distance between the input Gaussian Model and the closest trained Gaussian Model;
- classifying at least the portion as one of music, silence, or environment sound if the distance is greater than a first threshold value;
- determining an energy distribution of the plurality of frames in a first bandwidth;
- and
- classifying at least the portion as one of music, silence, or environment sound if the distance is greater than a second threshold value and the energy distribution of the plurality of frames in the first bandwidth is less than a third threshold value, wherein the second threshold value is less than the first threshold value.
Dependencies are inferred from the claim text.