US 9,812,149 B2
Methods and systems for providing consistency in noise reduction during speech and non-speech periods
Kuan-Chieh Yen, Foster City, CA (US)
Assigned to Knowles Electronics, LLC, Itasca, IL (US)
Filed by Knowles Electronics, LLC, Itasca, IL (US)
Filed on Jan. 28, 2016, as Appl. No. 15/9,740.
Prior Publication US 2017/0221501 A1, Aug. 3, 2017
This patent is subject to a terminal disclaimer.
Int. Cl. G10L 21/00 (2013.01); G10L 21/0216 (2013.01); G10L 25/21 (2013.01); G10L 25/93 (2013.01); G10L 21/0232 (2013.01); H04R 3/00 (2006.01)
CPC G10L 21/0216 (2013.01) [G10L 21/0232 (2013.01); G10L 25/21 (2013.01); G10L 25/93 (2013.01); H04R 3/005 (2013.01); G10L 2021/02166 (2013.01); G10L 2021/02168 (2013.01); G10L 2025/937 (2013.01)] 24 Claims
OG exemplary drawing
 
1. A method for audio processing, the method comprising:
receiving a first signal including at least a voice component and a second signal including at least the voice component modified by at least a human tissue of a user, the voice component being speech of the user, the first and second signals including periods when the speech of the user is not present;
assigning a first weight to the first signal and a second weight to the second signal;
processing the first signal to obtain a first power estimate;
processing the second signal to obtain a second power estimate;
utilizing the first and second power estimates to identify the periods when the speech of the user is not present;
for the periods that have been identified to be when the speech of the user is not present, performing one or both of decreasing the first weight and increasing the second weight so as to enhance the level of the second signal relative to the first signal;
blending, based on the first weight and the second weight, the first signal and the second signal to generate an enhanced voice signal; and
prior to the assigning, aligning the second signal with the first signal, the aligning including applying a spectral alignment filter to the second signal.