US 12,254,895 B2
Detecting and compensating for the presence of a speaker mask in a speech signal
Thomas Clark, Westford, MA (US); and John C. Hardwick, Acton, MA (US)
Assigned to Digital Voice Systems, Inc., Westford, MA (US)
Filed by Digital Voice Systems, Inc., Westford, MA (US)
Filed on Jul. 2, 2021, as Appl. No. 17/366,782.
Prior Publication US 2023/0005498 A1, Jan. 5, 2023
Int. Cl. G10L 21/00 (2013.01); G10L 21/0272 (2013.01); G10L 25/18 (2013.01); G10L 25/78 (2013.01)
CPC G10L 25/78 (2013.01) [G10L 21/0272 (2013.01); G10L 25/18 (2013.01)] 23 Claims
OG exemplary drawing
 
1. A method of compensating a speech signal for the presence of a speaker mask, the method comprising:
receiving a speech signal representing speech of a speaker;
dividing the speech signal into subframes;
generating speech parameters for a subframe;
using the speech parameters for the subframe in determining whether the subframe is suitable for use in detecting a mask worn by the speaker;
upon determining that the subframe is suitable for use in detecting a mask, using the speech parameters for the subframe in determining whether a mask is present; and
upon determining that a mask is present, modifying the speech parameters for the subframe to produce modified speech parameters that compensate the speech signal for the presence of the mask.