Google Inc.

Other

Pedro Jose Moreno Mengibar

Petar Aleksic

Fadi Biadsy

Bhuvana Ramabhadran

Eugene Weinstein

Andrew M Rosenberg

Michael H Cohen

Mark Edward Epstein

Alexander Houston Gruenstein

Johan Schalkwyk

Yaniv Leviathan

Rohit Prakash Prabhavalkar

Arun Narayanan

Yu Zhang

Ron J Weiss

Michiel A U Bacchiani

Tara N Sainath

Ignacio Lopez Moreno

Diego Melendo Casado

Yonghui Wu

Bo Li

Zhifeng Chen

Patrick An Phu Nguyen

Khe Chai Sim

Shumeet Baluja

Francoise Beaufays

Brian Strope

Fuchun Peng

Jeffrey Scott Sorensen

Ehsan Variani

Daniel Martin Bikel

Quan Wang

Michael Dennis Riley

Sangsoo Sung

Trevor D Strohman

Aleksandar Kracun

Françoise Beaufays

Zhehuai Chen

Neeraj Gaur

Parisa Haghani

Tomer Amiaz

Asaf Aharoni

Austin Severn Waters

Nir Shabat

Zhongdi Qu

Galen Tsai Chuang

Olivier Siohan

Tongzhou Chen

Leonid Velikovich

Kaisuke Nakajima

Olga Kapralova

Martin Jansche

Assaf Hurwitz Michaely

Kartik Audhkhasi

John Paul Alex

Jin Shi

Xin Lei

Yuli Gao

Jason William Pelecanos

Brian E Roark

Gang Feng

Jesse Emond

Fang Chu

Yuan Wang

Min Ma

Isabel Leal

Brian Farris

Manasa Prasad

Liyang Jiang

Yun Zhu

Yeming Fang

Michiel Au Bacchiani

Gary Wang

Petar Alecsic

Brian Patrick Strope

A method includes receiving training data that includes unspoken textual utterances, un-transcribed non-synthetic speech utterances, and transcribed non-synthetic speech utterances. Each unspoken textual utterance is not paired with any corresponding spoken utterance of non-synthetic speech. Each un-transcribed non-synthetic speech utterance not paired with a corresponding transcription. Each transcribed non-synthetic speech utterance paired with a corresponding transcription. The method also includes generating a corresponding alignment output for each unspoken textual utterance of the received training data using an alignment model. The method also includes pre-training an audio encoder on the alignment outputs generated for corresponding to the unspoken textual utterances, the un-transcribed non-synthetic speech utterances, and the transcribed non-synthetic speech utterances to teach the audio encoder to jointly learn shared speech and text representations.

Alignment prediction to inject text into automatic speech recognition training

A method for distilling one or more trained teacher automatic speech recognition (ASR) models into a multilingual student model includes receiving a plurality of teacher training examples and a plurality of student training examples. The method also includes training one or more teacher automatic speech recognition (ASR) models using the plurality of teacher training examples. Each teacher ASR model is configured to output a respective textual representation of a respective audio input. The method further includes generating a multi-lingual student ASR model by training the multi-lingual student ASR model using the plurality of student training examples and distilling the trained one or more teacher ASR models into the multilingual student ASR model using a tunable distillation loss weight. Each student ASR model is configured to receive an audio input and output a corresponding textual representation of the received audio input.

Self-adaptive distillation

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for cross-lingual speech recognition are disclosed. In one aspect, a method includes the actions of determining a context of a second computing device. The actions further include identifying, by a first computing device, an additional pronunciation for a term of multiple terms. The actions further include including the additional pronunciation for the term in the lexicon. The actions further include receiving audio data of an utterance. The actions further include generating a transcription of the utterance by using the lexicon that includes the multiple terms and the pronunciation for each of the multiple terms and the additional pronunciation for the term. The actions further include after generating the transcription of the utterance, removing the additional pronunciation for the term from the lexicon. The actions further include providing, for output, the transcription.

Cross-lingual speech recognition

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for implementing contextual grammar selection are disclosed. In one aspect, a method includes the actions of receiving audio data of an utterance. The actions include generating a word lattice that includes multiple candidate transcriptions of the utterance and that includes transcription confidence scores. The actions include determining a context of the computing device. The actions include based on the context of the computing device, identifying grammars that correspond to the multiple candidate transcriptions. The actions include determining, for each of the multiple candidate transcriptions, grammar confidence scores that reflect a likelihood that a respective grammar is a match for a respective candidate transcription. The actions include selecting, from among the candidate transcriptions, a candidate transcription. The actions further include providing, for output, the selected candidate transcription as a transcription of the utterance.

Contextual tagging and biasing of grammars inside word lattices

A method for training a sub-model for contextual biasing for speech recognition includes obtaining a base speech recognition model trained on non-biased data. The method includes obtaining a set of training utterances representative of a particular domain, each training utterance in the set of training utterances including audio data characterizing the training utterances and a ground truth transcription of the training utterance. The method further includes, for each corresponding training utterance in the set of training utterances, determining, using an embedding encoder, a corresponding document embedding from the ground truth transcription of the corresponding training utterance. The method includes training, using the corresponding document embeddings determined from the ground truth transcriptions of the set of training utterances, a sub-model to bias the base speech recognition model to recognize speech in the particular domain.

Sub-models for neural contextual biasing with attention and embedding space

Implementations set forth herein relate to an automated assistant that is invoked according to contextual signals—in lieu of requiring a user to explicitly speak an invocation phrase. When a user is in an environment with an assistant-enabled device, contextual data characterizing features of the environment can be processed to determine whether a user intends to invoke the automated assistant. Therefore, when such features are detected by the automated assistant, the automated assistant can bypass requiring an invocation phrase from a user and, instead, be responsive to one or more assistant commands from the user. The automated assistant can operate based on a trained machine learning model that is trained using instances of training data that characterize previous interactions in which one or more users invoked or did not invoke the automated assistant.

Selectively invoking an automated assistant based on detected environmental conditions without necessitating voice-based invocation of the automated assistant

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for end to end speech conversion are disclosed. In one aspect, a method includes the actions of receiving first audio data of a first utterance of one or more first terms spoken by a user. The actions further include providing the first audio data as an input to a model that is configured to receive first given audio data in a first voice and output second given audio data in a synthesized voice without performing speech recognition on the first given audio data. The actions further include receiving second audio data of a second utterance of the one or more first terms spoken in the synthesized voice. The actions further include providing, for output, the second audio data of the second utterance of the one or more first terms spoken in the synthesized voice.

End-to-end speech conversion

A method includes receiving training data that includes unspoken text utterances, un-transcribed non-synthetic speech utterances, and transcribed non-synthetic speech utterances. Each unspoken text utterance is not paired with any corresponding spoken utterance of non-synthetic speech. Each un-transcribed non-synthetic speech utterance is not paired with a corresponding transcription. Each transcribed non-synthetic speech utterance is paired with a corresponding transcription. The method also includes generating a corresponding synthetic speech representation for each unspoken textual utterance of the received training data using a text-to-speech model. The method also includes pre-training an audio encoder on the synthetic speech representations generated for the unspoken textual utterances, the un-transcribed non-synthetic speech utterances, and the transcribed non-synthetic speech utterances to teach the audio encoder to jointly learn shared speech and text representations.

Advancing the use of text and speech in ASR pretraining with consistency and contrastive losses

A method for speech conversion includes receiving, as input to an encoder of a speech conversion model, an input spectrogram corresponding to an utterance, the encoder including a stack of self-attention blocks. The method further includes generating, as output from the encoder, an encoded spectrogram and receiving, as input to a spectrogram decoder of the speech conversion model, the encoded spectrogram generated as output from the encoder. The method further includes generating, as output from the spectrogram decoder, an output spectrogram corresponding to a synthesized speech representation of the utterance.

Conformer-based speech conversion model

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for an automated calling system are disclosed. In one aspect, a method includes the actions of receiving audio data of an utterance spoken by a user who is having a telephone conversation with a bot. The actions further include determining a context of the telephone conversation. The actions further include determining a user intent of a first previous portion of the telephone conversation spoken by the user and a bot intent of a second previous portion of the telephone conversation outputted by a speech synthesizer of the bot. The actions further include, based on the audio data of the utterance, the context of the telephone conversation, the user intent, and the bot intent, generating synthesized speech of a reply by the bot to the utterance. The actions further include, providing, for output, the synthesized speech.

Automated calling system

Methods, systems, and apparatus, including computer programs encoded on a computer-readable media, for speech recognition using multi-dialect and multilingual models. In some implementations, audio data indicating audio characteristics of an utterance is received. Input features determined based on the audio data are provided to a speech recognition model that has been trained to output score indicating the likelihood of linguistic units for each of multiple different language or dialects. The speech recognition model can be one that has been trained using cluster adaptive training. Output that the speech recognition model generated in response to receiving the input features determined based on the audio data is received. A transcription of the utterance generated based on the output of the speech recognition model is provided.

Multi-dialect and multilingual speech recognition

A method includes receiving a sequence of acoustic frames extracted from audio data corresponding to an utterance. During a first pass, the method includes processing the sequence of acoustic frames to generate N candidate hypotheses for the utterance. During a second pass, and for each candidate hypothesis, the method includes: generating a respective un-normalized likelihood score; generating a respective external language model score; generating a standalone score that models prior statistics of the corresponding candidate hypothesis; and generating a respective overall score for the candidate hypothesis based on the un-normalized likelihood score, the external language model score, and the standalone score. The method also includes selecting the candidate hypothesis having the highest respective overall score from among the N candidate hypotheses as a final transcription of the utterance.

Multilingual re-scoring models for automatic speech recognition

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for modulating language model biasing. In some implementations, context data is received. A likely context associated with a user is determined based on at least a portion of the context data. One or more language model biasing parameters based at least on the likely context associated with the user is selected. A context confidence score associated with the likely context based on at least a portion of the context data is determined. One or more language model biasing parameters based at least on the context confidence score is adjusted. A baseline language model based at least on the one or more of the adjusted language model biasing parameters is biased. The baseline language model is provided for use by an automated speech recognizer (ASR).

Language model biasing modulation

A method for contextual biasing for speech recognition includes obtaining a base automatic speech recognition (ASR) model trained on non-biased data and a sub-model trained on biased data representative of a particular domain. The method includes receiving a speech recognition request including audio data characterizing an utterance captured in streaming audio. The method further includes determining whether the speech recognition request includes a contextual indicator indicating the particular domain. When the speech recognition request does not include the contextual indicator, the method includes generating, using the base ASR model, a first speech recognition result of the utterance by processing the audio data. When the speech recognition request includes the contextual indicator the method includes biasing, using the sub-model, the base ASR model toward the particular domain and generating, using the biased base ASR model, a second speech recognition result of the utterance by processing the audio data.

Sub-models for neural contextual biasing

Systems, methods, devices, and other techniques are described herein for determining dialog states that correspond to voice inputs and for biasing a language model based on the determined dialog states. In some implementations, a method includes receiving, at a computing system, audio data that indicates a voice input and determining a particular dialog state, from among a plurality of dialog states, which corresponds to the voice input. A set of n-grams can be identified that are associated with the particular dialog state that corresponds to the voice input. In response to identifying the set of n-grams that are associated with the particular dialog state that corresponds to the voice input, a language model can be biased by adjusting probability scores that the language model indicates for n-grams in the set of n-grams. The voice input can be transcribed using the adjusted language model.

Determining dialog states for language models

Growing community of inventors

Jersey City, NJ, United States of America

Pedro Jose Moreno Mengibar