Noise reduction systems and methods for voice applications
First Claim
Patent Images
1. A method comprising:
- providing a computing device having an array of microphones comprising one or more microphones; and
using the microphone array, training the device to recognize noise from known locations by equipping the device with a filter system that can filter noise from the known locations.
2 Assignments
0 Petitions
Accused Products
Abstract
Various embodiments reduce noise within a particular environment, while isolating and capturing speech in a manner that allows operation within an otherwise noisy environment. In one embodiment, an array of one or more microphones is used to selectively eliminate noise emanating from known, generally fixed locations, and pass signals from a pre-specified region or regions with reduced distortion.
-
Citations
111 Claims
-
1. A method comprising:
-
providing a computing device having an array of microphones comprising one or more microphones; and
using the microphone array, training the device to recognize noise from known locations by equipping the device with a filter system that can filter noise from the known locations. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10)
-
-
11. A method comprising:
-
providing a computing device having an array of microphones comprising one or more microphones; and
using the microphone array, training the device to recognize noise from particular known sources by equipping the device with a filter system that can filter noise from the particular known sources. - View Dependent Claims (12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25)
-
-
26. A method comprising:
-
providing a game controller having an array of microphones comprising one or more microphones;
using the microphone array, training the game controller to recognize audio signals from particular known locations and sources by equipping the game controller with a filter system that can (a) filter noise from particular known locations and sources, and (b) pass signals associated with desired speech from particular locations. - View Dependent Claims (27, 28, 29, 30, 31, 32, 33, 34, 35)
-
-
36. A method comprising:
-
providing a user-engagable input device comprising a housing that supports an array of microphones, at least one of the microphones being mounted inside of the housing;
using the microphone array, capturing audio signals associated with noise;
correlation processing the audio signals associated with the noise and constructing one or more filter components as a function of the processed audio signals;
using the microphone array, capturing audio signals associated with desired speech;
correlation processing the audio signals associated with desired speech and constructing one or more filter components as a function of the processed audio speech signals; and
incorporating a filter system comprising the filter components into one or more user engagable input devices. - View Dependent Claims (37, 38, 39, 40)
-
-
41. A method comprising:
-
providing a computing device having an array of microphones comprising one or more microphones, the computing device comprising a trained filter system configured to recognize noise from particular known locations relative to the computing device;
capturing audio signals using the microphone array;
filtering noise from the captured audio signals using the trained filter system. - View Dependent Claims (42, 43, 44, 45, 46, 47, 48, 49, 50, 51)
-
-
52. A method comprising:
-
providing a computing device having an array of microphones comprising one or more microphones, the computing device comprising a trained filter system configured to recognize noise from particular known sources;
capturing audio signals using the microphone array; and
filtering noise from the captured audio signals using the trained filter system. - View Dependent Claims (53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66)
-
-
67. A method comprising:
-
providing a game controller having an array of microphones comprising one or more microphones, the game controller comprising a trained filter system configured to recognize audio signals from particular known locations and sources;
capturing audio signals using the microphone array;
filtering the captured signals using the trained filter system effective to (a) filter noise from particular locations and sources, and (b) pass signals associated with desired speech from particular locations. - View Dependent Claims (68, 69, 70, 71, 72, 73, 74, 75)
-
-
76. A method comprising:
-
providing a user-engagable input device comprising a housing that supports an array of microphones, at least one of the microphones being mounted inside of the housing;
capturing audio signals associated with the environment in which the user-engagable input device is used, wherein the audio signals can comprise both noise and desired speech;
filtering the captured audio signals using a trained filter system that is configured to recognize noise and desired speech, the filter system comprising multiple filters computed as generalized Wiener filters having the form;
wopt=(Rss+β
Rnn)−
1(E{ds}),where Rss is the correlation matrix for a desired speech signal, Rnn is the correlation matrix for the noise component, β
is a weighting parameter for the noise component, and E{ds} is the expected value of the product of the desired signal d and the actual signal s that is received by a microphone. - View Dependent Claims (77, 78, 79, 80, 81)
-
-
82. A system comprising:
-
a housing;
one or more user input mechanisms supported by the housing;
a processor;
a computer-readable media;
a microphone array at least some of which supported by the housing and comprising one or more microphones;
a noise reduction component comprising a filter system embodied on the computer-readable media, the filter system being trained to recognize noise from particular known locations; and
the noise reduction component being configured to cause the processor to use the trained filter system to filter noise, from said known locations, from audio signals captured by the microphone array. - View Dependent Claims (83, 84, 85, 86, 87, 88, 89)
-
-
90. A system comprising:
-
a housing;
one or more user input mechanisms supported by the housing;
a processor;
a computer-readable media;
a microphone array comprising one or more microphones;
a noise reduction component comprising a filter system embodied on the computer-readable media, the filter system being trained to recognize noise from particular known sources; and
the noise reduction component being configured to cause the processor to use the trained filter system to filter noise, from said known sources, from audio signals captured by the microphone array. - View Dependent Claims (91, 92, 93, 94, 95)
-
-
96. A system comprising:
-
a housing;
one or more user input mechanisms supported by the housing;
a processor;
a computer-readable media;
a microphone array comprising one or more microphones, at least one of the microphones being mounted within the housing;
a noise reduction component comprising a filter system embodied on the computer-readable media, the filter system being trained to recognize audio signals from particular known sources and locations; and
the noise reduction component being configured to cause the processor to use the trained filter system to (a) filter noise, from said known sources and locations, from audio signals captured by the microphone array, and (b) pass signals associated with desired speech from particular locations. - View Dependent Claims (97, 98, 99, 100, 101, 102)
-
-
103. A noise reduction component comprising:
-
a transform component configured to transform audio samples from a microphone array from the time domain into the frequency domain;
a filter system associated with the transform component and configured to filter frequency samples produced by the transform component, the filter system comprising multiple filters each of which being associated with a frequency bin, individual filters comprising a generalized Wiener filter having the form;
wopt=(Rss+β
Rnn)−
1(E{ds}),where Rss is the correlation matrix for a desired speech signal, Rnn is the correlation matrix for a noise component, β
is a weighting parameter for the noise component, and E{ds} is the expected value of the product of the desired signal d and the actual signal s that is received by a microphone. - View Dependent Claims (104, 105, 106, 107, 108, 109, 110, 111)
-
Specification