| CPC G10L 21/0216 (2013.01) [G10L 25/18 (2013.01)] | 20 Claims |

|
1. A method comprising:
receiving an audio signal;
determining whether a current portion of the audio signal comprises speech or non-speech;
extracting from the audio signal a plurality of time-frequency features, the plurality of time-frequency features comprising one or more short-term features;
determining whether an impulsive sound is present in the extracted one or more short-term features;
in accordance with determining that non-speech is present in the current portion of the audio signal, adapting one or more thresholds of a sibilance detector to a first set of one or more threshold values, the adapting including giving a first weight to a first output value resulting from the determination of whether the impulsive sound is present;
in accordance with determining that speech is present in the current portion of the audio signal, adapting the one or more thresholds of the sibilance detector to a second set of one or more threshold values different than the first set of one or more threshold values, the adapting including giving a second weight to the first output value resulting from the determination of whether the impulsive sound is present;
detecting sibilance in the audio signal, using the sibilance detector with the one or more adapted thresholds; and
suppressing the sibilance in the audio signal by applying to the audio signal one or more gains determined in response to the sibilance in the audio signal.
|