US 12,243,545 B2
Method and system of neural network dynamic noise suppression for audio processing
Adam Kupryjanow, Gdansk (PL); and Lukasz Pindor, Gdansk (PL)
Assigned to Intel Corporation, Santa Clara, CA (US)
Filed by Intel Corporation, Santa Clara, CA (US)
Filed on Dec. 24, 2021, as Appl. No. 17/561,848.
Claims priority of provisional application 63/252,687, filed on Oct. 6, 2021.
Prior Publication US 2022/0124433 A1, Apr. 21, 2022
Int. Cl. H03G 5/00 (2006.01); G06N 3/08 (2023.01); G10L 21/0208 (2013.01); G10L 21/0232 (2013.01); G10L 25/30 (2013.01); G10L 25/78 (2013.01); H04R 3/04 (2006.01)
CPC G10L 21/0208 (2013.01) [G06N 3/08 (2013.01); G10L 21/0232 (2013.01); G10L 25/30 (2013.01); G10L 25/78 (2013.01); H04R 3/04 (2013.01)] 25 Claims
OG exemplary drawing
 
1. A computer-implemented method of audio processing, comprising:
obtaining audio signal data from a noise suppression encoder receiving an audio signal;
inputting an input vector of multiple channels associated with the audio signal data into one or more grouped pointwise convolutional layers of a neural network of a noise suppression separator,
wherein each grouped pointwise convolutional layer being trained to have at least two channels of the input vector share the same one or more neural network weight values; and
outputting a noise suppression mask from the neural network to apply to a version of the audio signal.