Granted Patent B1

US 12,401,942 · App. 18/323,697 · Granted Aug 26, 2025

Inventors: Robert Ayrapetian (Morgan Hill, CA); Gautam Shreedhar Bhat (Sunnyvale, CA); Pradeep Kumar Govindaraju (San Jose, CA)

Assignee: Amazon Technologies, Inc.

H04R1/32H04B11/00

Abstract

Claims (112)

1. A computer-implemented method, the method comprising:

receiving first audio data including a first portion corresponding to a first direction and a second portion corresponding to a second direction, the first audio data including a first representation of speech;

determining, using the first portion of the first audio data, a first value representing a first noise floor of the first portion of the first audio data;

determining, using the second portion of the first audio data, a second value representing a second noise floor of the second portion of the first audio data;

determining, using the first value, a first signal quality metric value corresponding to the first portion of the first audio data;

determining, using the first value and the second value, a first ratio value; and

determining, using the first signal quality metric value and the first ratio value, second audio data corresponding to a first plurality of directions including the first direction, the second audio data including a second representation of the speech.

2. The computer-implemented method of claim 1 , wherein determining the second audio data further comprises:

determining, using a third portion of the first audio data corresponding to a third direction, a third value representing a third noise floor of the third portion;

determining, using the third value, a second signal quality metric value corresponding to the third portion;

determining, using the third value and the second value, a second ratio value;

determining a first portion of the second audio data using the first portion of the first audio data, the first signal quality metric value, the first ratio value, the second signal quality metric value, and the second ratio value; and

determining a second portion of the second audio data using the third portion of the first audio data, the first signal quality metric value, the first ratio value, the second signal quality metric value, and the second ratio value.

3. The computer-implemented method of claim 1 , wherein determining the second audio data further comprises:

determining a first product of the first ratio value and the first signal quality metric value;

determining a second product of a second ratio value and a second signal quality metric value, the second ratio value and the second signal quality metric value associated with a third portion of the first audio data corresponding to a third direction;

determining a first weight value using the first product and a sum of the first product and the second product;

determining a second weight value using the second product and the sum of the first product and the second product;

determining, using the first weight value and the first portion of the first audio data, a first portion of the second audio data; and

determining, using the second weight value and the third portion of the first audio data, a second portion of the second audio data.

4. The computer-implemented method of claim 1 , wherein determining the second audio data further comprises:

determining a first group value using the first signal quality metric value and the first ratio value, the first group value corresponding to the first plurality of directions;

determining that the first group value satisfies a condition; and

determining, using the first plurality of directions and the first audio data, the second audio data.

5. The computer-implemented method of claim 1 , wherein determining the second audio data further comprises:

determining a first group value using the first signal quality metric value and the first ratio value, the first group value corresponding to the first plurality of directions;

determining a second group value corresponding to a second plurality of directions including the second direction;

determining that the first group value is higher than the second group value; and

determining, using the first plurality of directions and the first audio data, the second audio data.

6. The computer-implemented method of claim 1 , further comprising:

determining, using a third portion of the first audio data corresponding to a third direction, a third value representing a third noise floor of the third portion of the first audio data;

determining that the second value is greater than the first value and the third value;

determining, using the second value, a second ratio value corresponding to the second portion of the first audio data; and

determining, using the third value and the second value, a third ratio value corresponding to the third portion of the first audio data.

7. The computer-implemented method of claim 6 , wherein determining the second audio data further comprises:

determining a first product of the first ratio value and the first signal quality metric value;

determining a second product of the third ratio value and a second signal quality metric value associated with the third portion of the first audio data;

determining that a sum of the first product and the second product satisfies a condition; and

determining, using the first plurality of directions and the first audio data, the second audio data, wherein the first plurality of directions includes the first direction and the third direction.

8. The computer-implemented method of claim 6 , wherein determining the second audio data further comprises:

determining a first product of the first ratio value and the first signal quality metric value;

determining a second product of the third ratio value and a second signal quality metric value associated with the third portion of the first audio data;

determining, using the first product and the first portion of the first audio data, a first portion of the second audio data; and

determining, using the second product and the third portion of the first audio data, a second portion of the second audio data.

9. The computer-implemented method of claim 1 , wherein determining the second audio data further comprises:

determining a first group value using the first signal quality metric value and the first ratio value;

determining a second group value corresponding to a second plurality of directions that includes the second direction;

determining a third group value using the first group value and a first weight value associated with the first plurality of directions;

determining a fourth group value using the second group value and a second weight value associated with the second plurality of directions;

determining that the third group value is higher than the fourth group value; and

determining, using the first plurality of directions and the first audio data, the second audio data.

10. The computer-implemented method of claim 1 , wherein determining the first value further comprises:

determining, using the first portion of the first audio data, first power values;

determining, using the first power values, first noise floor data; and

determining the first value using a lowest value represented in the first noise floor data.

11. A system comprising:

at least one processor; and

memory including instructions operable to be executed by the at least one processor to cause the system to:

receive first audio data including a first portion corresponding to a first direction and a second portion corresponding to a second direction, the first audio data including a first representation of speech;

determine, using the first portion of the first audio data, a first value representing a first noise floor of the first portion of the first audio data;

determine, using the second portion of the first audio data, a second value representing a second noise floor of the second portion of the first audio data;

determine, using the first value, a first signal quality metric value corresponding to the first portion of the first audio data;

determine, using the first value and the second value, a first ratio value; and

determine, using the first signal quality metric value and the first ratio value, second audio data corresponding to a first plurality of directions including the first direction, the second audio data including a second representation of the speech.

12. The system of claim 11 , wherein the memory further comprises instructions that, when executed by the at least one processor, further cause the system to:

determine, using a third portion of the first audio data corresponding to a third direction, a third value representing a third noise floor of the third portion;

determine, using the third value, a second signal quality metric value corresponding to the third portion;

determine, using the third value and the second value, a second ratio value;

determine a first portion of the second audio data using the first portion of the first audio data, the first signal quality metric value, the first ratio value, the second signal quality metric value, and the second ratio value; and

determine a second portion of the second audio data using the third portion of the first audio data, the first signal quality metric value, the first ratio value, the second signal quality metric value, and the second ratio value.

13. The system of claim 11 , wherein the memory further comprises instructions that, when executed by the at least one processor, further cause the system to:

determine a first product of the first ratio value and the first signal quality metric value;

determine a second product of a second ratio value and a second signal quality metric value, the second ratio value and the second signal quality metric value associated with a third portion of the first audio data corresponding to a third direction;

determine a first weight value using the first product and a sum of the first product and the second product;

determine a second weight value using the second product and the sum of the first product and the second product;

determine, using the first weight value and the first portion of the first audio data, a first portion of the second audio data; and

determine, using the second weight value and the third portion of the first audio data, a second portion of the second audio data.

14. The system of claim 11 , wherein the memory further comprises instructions that, when executed by the at least one processor, further cause the system to:

determine a first group value using the first signal quality metric value and the first ratio value, the first group value corresponding to the first plurality of directions;

determine that the first group value satisfies a condition; and

determine, using the first plurality of directions and the first audio data, the second audio data.

15. The system of claim 11 , wherein the memory further comprises instructions that, when executed by the at least one processor, further cause the system to:

determine a first group value using the first signal quality metric value and the first ratio value, the first group value corresponding to the first plurality of directions;

determine a second group value corresponding to a second plurality of directions including the second direction;

determine that the first group value is higher than the second group value; and

determine, using the first plurality of directions and the first audio data, the second audio data.

16. The system of claim 11 , wherein the memory further comprises instructions that, when executed by the at least one processor, further cause the system to:

determine, using a third portion of the first audio data corresponding to a third direction, a third value representing a third noise floor of the third portion of the first audio data;

determine that the second value is greater than the first value and the third value;

determine, using the second value, a second ratio value corresponding to the second portion of the first audio data; and

determine, using the third value and the second value, a third ratio value corresponding to the third portion of the first audio data.

17. The system of claim 16 , wherein the memory further comprises instructions that, when executed by the at least one processor, further cause the system to:

determine a first product of the first ratio value and the first signal quality metric value;

determine a second product of the third ratio value and a second signal quality metric value associated with the third portion of the first audio data;

determine that a sum of the first product and the second product satisfies a condition; and

determine, using the first plurality of directions and the first audio data, the second audio data, wherein the first plurality of directions includes the first direction and the third direction.

18. The system of claim 16 , wherein the memory further comprises instructions that, when executed by the at least one processor, further cause the system to:

determine a first product of the first ratio value and the first signal quality metric value;

determine a second product of the third ratio value and a second signal quality metric value associated with the third portion of the first audio data;

determine, using the first product and the first portion of the first audio data, a first portion of the second audio data; and

determine, using the second product and the third portion of the first audio data, a second portion of the second audio data.

19. The system of claim 11 , wherein the memory further comprises instructions that, when executed by the at least one processor, further cause the system to:

determine a first group value using the first signal quality metric value and the first ratio value;

determine a second group value corresponding to a second plurality of directions that includes the second direction;

determine a third group value using the first group value and a first weight value associated with the first plurality of directions;

determine a fourth group value using the second group value and a second weight value associated with the second plurality of directions;

determine that the third group value is higher than the fourth group value; and

determine, using the first plurality of directions and the first audio data, the second audio data.

20. The system of claim 11 , wherein the memory further comprises instructions that, when executed by the at least one processor, further cause the system to:

determine, using the first portion of the first audio data, first power values;

determine, using the first power values, first noise floor data; and

determine the first value using a lowest value represented in the first noise floor data.

Assignments (1)

References Cited (66)