Multi-microphone speech recognition systems and related techniques
First Claim
1. A system for resolving utterances, the system comprising a processor and a memory, wherein the memory contains instructions that, when executed by the processor, cause the system to:
- receive a plurality of utterance representations corresponding to an utterance, wherein each utterance representation corresponds to a respective version of the utterance observed by an acoustic device in a plurality of acoustic devices exposed to the utterance;
jointly process the plurality of utterance representations to determine recognized speech corresponding to the utterance; and
coordinate control of one or more of the plurality of acoustic devices responsive to the recognized speech.
1 Assignment
0 Petitions
Accused Products
Abstract
A speech recognition system for resolving impaired utterances can have a speech recognition engine configured to receive a plurality of representations of an utterance and concurrently to determine a plurality of highest-likelihood transcription candidates corresponding to each respective representation of the utterance. The recognition system can also have a selector configured to determine a most-likely accurate transcription from among the transcription candidates. As but one example, the plurality of representations of the utterance can be acquired by a microphone array, and beamforming techniques can generate independent streams of the utterance across various look directions using output from the microphone array.
52 Citations
33 Claims
-
1. A system for resolving utterances, the system comprising a processor and a memory, wherein the memory contains instructions that, when executed by the processor, cause the system to:
-
receive a plurality of utterance representations corresponding to an utterance, wherein each utterance representation corresponds to a respective version of the utterance observed by an acoustic device in a plurality of acoustic devices exposed to the utterance; jointly process the plurality of utterance representations to determine recognized speech corresponding to the utterance; and coordinate control of one or more of the plurality of acoustic devices responsive to the recognized speech. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16)
-
-
17. An acoustic appliance comprising a processor and a memory, wherein the memory stores instructions which, when executed by the processor, cause the acoustic appliance to, in real-time:
-
receive a plurality of versions of an utterance from a corresponding plurality of spatially distributed acoustic devices; process each version of the utterance to determine an utterance representation corresponding to each version of the utterance; over a communication connection, transmit the plurality of utterance representations to a computing component; and over the communication connection, receive from the computing component, recognized speech corresponding to the utterance. - View Dependent Claims (18, 19, 20, 21, 22, 23, 24, 25, 26)
-
-
27. An electronic device comprising a processor, a memory, and a plurality of microphone transducers, wherein the memory contains instructions that, when executed by the processor, cause the electronic device to, in real-time:
-
observe an utterance with each of the plurality of microphone transducers and determine an observed version of the utterance for each microphone transducer; determine a plurality of utterance representations from the plurality of observed versions of the utterance, wherein each utterance representation corresponds to a respective observed version of the utterance; communicate the plurality of utterance representations to a speech-recognition component/module over a communication connection; receive over the communication connection recognized speech corresponding to the utterance; and invoke a task responsive to the recognized speech. - View Dependent Claims (28, 29, 30, 31, 32, 33)
-
Specification