-
公开(公告)号：US20190156836A1
公开(公告)日：2019-05-23
申请号：US16192914
申请日：2018-11-16
Inventor： Rahim SAEIDI
Abstract: A method of speaker modelling for a speaker recognition system, comprises: receiving a signal comprising a speaker's speech; and, for a plurality of frames of the signal: obtaining a spectrum of the speaker's speech; generating at least one modified spectrum, by applying effects related to a respective vocal effort; and extracting features from the spectrum of the speaker's speech and the at least one modified spectrum. The method further comprises forming at least one speech model based on the extracted features.
-
公开(公告)号：US20210219055A1
公开(公告)日：2021-07-15
申请号：US17057516
申请日：2019-05-29
Inventor： Pablo PESO PARADA , Rahim SAEIDI , John L. MELANSON
Abstract: The application describes techniques for howling detection. A howling detector is described that is configured to receive an input signal and to determine measure of the linearity of a logarithmic representation of the energy of the input signal. In some examples, this triggers gain adjustment (e.g. of a noise control unit) and, in some further examples, the amount of the gain adjustment may be based on an estimation of the maximum stable gain of a noise control unit.
-
公开(公告)号：US20190379972A1
公开(公告)日：2019-12-12
申请号：US16004900
申请日：2018-06-11
Inventor： Pablo PESO PARADA , Rahim SAEIDI , John L. MELANSON
Abstract: The application describes techniques for howling detection. A processing module for a noise control circuit is described. The processing module comprises a howling detector configured to receive an input signal and to determine a log-linearity metric based on the input signal. Gain adjustment of the noise control unit following the detection of howling is also described, wherein the amount of gain adjustment may be based on an estimation of the maximum stable gain of the noise control unit.
-
-
Search Results
3
records
(took 0.02 seconds)
