Abstract:
The subject matter of this specification can be embodied in, among other things, a method that includes receiving audio data that corresponds to an utterance, obtaining a first transcription of the utterance that was generated using a limited speech recognizer. The limited speech recognizer includes a speech recognizer that includes a language model that is trained over a limited speech recognition vocabulary that includes one or more terms from a voice command grammar, but that includes fewer than all terms of an expanded grammar. A second transcription of the utterance is obtained that was generated using an expanded speech recognizer. The expanded speech recognizer includes a speech recognizer that includes a language model that is trained over an expanded speech recognition vocabulary that includes all of the terms of the expanded grammar. The utterance is classified based at least on a portion of the first transcription or the second transcription.
Abstract:
A method includes accessing data specifying a set of actions, each action defining a user device operation and for each action: accessing a corresponding set of command sentences for the action, determining first n-grams in the set of command sentences that are semantically relevant for the action, determining second n-grams in the set of command sentences that are semantically irrelevant for the action, generating a training set of command sentences from the corresponding set of command sentences, the generating the training set of command sentences including removing each second n-gram from each sentence in the corresponding set of command sentences for the action, and generating a command model from the training set of command sentences configured to generate an action score for the action for an input sentence based on: first n-grams for the action, and second n-grams for the action that are also second n-grams for all other actions.
Abstract:
Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating a set of training utterances. The methods, systems, and apparatus include actions of obtaining a target multi-dimensional distribution of characteristics in an initial set of candidate utterances and selecting a subset of the initial set of candidate utterances based on speech recognition confidence scores associated with the candidate utterances. Additional actions include selecting a particular candidate utterance from the subset of the initial set of utterances and determining that adding the particular candidate utterance to a set of training utterances reduces a divergence of a multi-dimensional distribution of the characteristics in the set of training utterances from the target multi-dimensional distribution. Further actions include adding the particular candidate utterance to the set of training utterances.
Abstract:
This specification describes technologies relating to system, methods, and articles for updating a speech recognition dictionary based on, at least in part, both search query and market data metrics. In general, one innovative aspect of the subject matter described in this specification can be embodied in a method comprising (i) identifying a candidate term for possible inclusion in a speech recognition dictionary, (ii) identifying at least one search query metric associated with the identified candidate term, (iii) identifying at least one market data metric associated with the identified candidate term, and (iv) generating a candidate term score for the identified candidate term based, at least in part, on a weighted combination of the at least one identified search query metric and the at least one identified market data metric.
Abstract:
A processing system receives an audio signal encoding a portion of an utterance. The processing system receives context information associated with the utterance, wherein the context information is not derived from the audio signal or any other audio signal. The processing system provides, as input to a neural network, data corresponding to the audio signal and the context information, and generates a transcription for the utterance based on at least an output of the neural network.
Abstract:
Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating a set of training utterances. The methods, systems, and apparatus include actions of obtaining a target multi-dimensional distribution of characteristics in an initial set of candidate utterances and selecting a subset of the initial set of candidate utterances based on speech recognition confidence scores associated with the candidate utterances. Additional actions include selecting a particular candidate utterance from the subset of the initial set of utterances and determining that adding the particular candidate utterance to a set of training utterances reduces a divergence of a multi-dimensional distribution of the characteristics in the set of training utterances from the target multi-dimensional distribution. Further actions include adding the particular candidate utterance to the set of training utterances.
Abstract:
Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training recognition canonical representations corresponding to named-entity phrases in a second natural language based on translating a set of allowable expressions with canonical representations from a first natural language, which may be generated by expanding a context-free grammar for the allowable expressions for the first natural language.
Abstract:
Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training recognition canonical representations corresponding to named-entity phrases in a second natural language based on translating a set of allowable expressions with canonical representations from a first natural language, which may be generated by expanding a context-free grammar for the allowable expressions for the first natural language.