Parallel recognition

One or more non-transitory machine-readable media storing instructions that, when executed by one or more computers, are operable to cause the one or more computers to perform operations comprising:

• obtaining speech data for a particular frame representing a portion of an utterance;
• providing the speech data to multiple processors that are configured to process the speech data in parallel, each of the multiple processors being assigned to evaluate a corresponding set of path segments in a network of states;
• before the multiple processors receive speech data for future frames that occur in the utterance after the particular frame and represent subsequent portions of the utterance, receiving, from the multiple processors, cost values that represent an extent to which path segments match one or more of the future frames not yet received by the multiple processors;
• communicating, among the multiple processors, the cost values that represent the extent to which path segments match the one or more of the future frames not yet received by the multiple processors;
• and
• using the communicated cost values to determine a transcription of at least a portion of the utterance.

A system comprising:

• one or more computers and one or more storage devices storing instructions that are operable, when executed by the one or more computers, to cause the one or more computers to perform operations comprising:
• obtaining speech data for a particular frame representing a portion of an utterance;
• providing the speech data to multiple processors that are configured to process the speech data in parallel, each of the multiple processors being assigned to evaluate a corresponding set of path segments in a network of states;
• before the multiple processors receive speech data for future frames that occur in the utterance after the particular frame and represent subsequent portions of the utterance, receiving, from the multiple processors, cost values that represent an extent to which path segments match one or more of the future frames not yet received by the multiple processors;
• communicating, among the multiple processors, the cost values that represent the extent to which path segments match the one or more of the future frames not yet received by the multiple processors;
• and
• using the communicated cost values to determine a transcription of at least a portion of the utterance.

A method comprising:

• obtaining, at multiple processors of different computers that are configured to cooperatively recognize speech by processing speech data in parallel, speech data indicating characteristics of a portion of an utterance that occurs during a particular time period;
• and
• before the multiple processors receive speech data for portions of the utterance that occur during future time periods that occur after the particular time period:
• determining, by each of the multiple processors operating in parallel, cost values representing an extent to which respective words match portions of the utterance occurring during the future time periods;
• communicating, by each of the multiple processors and over a computer network, one or more of the cost values determined by the processor;
• and
• communicating, by each of the multiple processors, additional cost values representing the extent to which the respective words match the portions of the utterance only after a predetermined amount of time comprising multiple of the future time periods has elapsed.

A system comprising:

• multiple processors that are physically distinct, the multiple processors being connected through a communication network;
• and
• storage devices storing instructions that are operable, when executed by the one or more processors, to cause the one or more processors to perform operations comprising:
• obtaining, at the multiple processors, speech data indicating characteristics of a portion of an utterance that occurs during a particular time period;
• and
• before the multiple processors receive speech data for portions of the utterance that occur during future time periods that occur after the particular time period:
• determining, by each of the multiple processors operating in parallel, cost values representing an extent to which respective words match portions of the utterance occurring during the future time periods;
• communicating, by each of the multiple processors and over a computer network, one or more of the cost values determined by the processor;
• and
• communicating, by each of the multiple processors, additional cost values representing the extent to which the respective words match the portions of the utterance only after a predetermined amount of time comprising multiple of the future time periods has elapsed.

One or more non-transitory machine-readable media storing instructions that, when executed by multiple processors, are operable to cause the multiple processors to perform operations comprising:

• obtaining, at the multiple processors, speech data indicating characteristics of a portion of an utterance that occurs during a particular time period;
• and
• before the multiple processors receive speech data for portions of the utterance that occur during future time periods that occur after the particular time period:
• determining, by each of the multiple processors operating in parallel, cost values representing an extent to which respective words match portions of the utterance occurring during the future time periods;
• communicating, by each of the multiple processors and over a computer network, one or more of the cost values determined by the processor;
• and
• communicating, by each of the multiple processors, additional cost values representing the extent to which the respective words match the portions of the utterance only after a predetermined amount of time comprising multiple of the future time periods has elapsed.