Google Inc.

Yonghui Wu

Zhifeng Chen

Ron J Weiss

Yu Zhang

Tara N Sainath

Ruoming Pang

Ye Jia

Jonathan Shen

Byungha Chun

Chung-Cheng Chiu

Patrick An Phu Nguyen

Paul Haahr

Navdeep Jaitly

Michael Schuster

Quoc V Le

Bo Li

Rohit Prakash Prabhavalkar

Ioannis Agiomyrgiannakis

Ryan Michael Rifkin

Michael E Flaster

Melvin Jose Johnson Premkumar

Anjuli Patricia Kannan

Yuxuan Wang

Zongheng Yang

Niki J Parmar

Russell John Wyatt Skerry-Ryan

Wei Han

Samuel Bengio

Wolfgang Macherey

Robert Andrew James Clark

Orhan Firat

Ankur Bapna

Russell J Ryan

Jiahui Yu

Xu Chen

Isaac Elias

Ying Xiao

Nitin Kumar Gupta

Pedro Jose Moreno Mengibar

Bhuvana Ramabhadran

Ignacio Lopez Moreno

Steven D Baker

Srinivasan Venkatachary

Quan Wang

Andrew M Rosenberg

Eugene Weinstein

Khe Chai Sim

Llion Owen Jones

Macduff Richard Hughes

Anmol Gulati

George Foster

Fei Ren

Maxim Krikun

Jeffrey Adgate Dean

Sanjay Ghemawat

Michael Isard

Chandramohan Amyangot Thekkath

Thorsten Brants

Michiel A U Bacchiani

Yanzhang He

Kanury Kanishka Rao

Matthew D Cutts

Paul Ronald Barham

Sudip Roy

Ryan Sepassi

Randall G Keller

Patrick An Nguyen

Brian Strope

Fadi Biadsy

William Chan

Timothy Youngjin Sohn

Trevor D Strohman

Arun Narayanan

Andrew M Dai

Mohammad Norouzi

Tobias Kaufmann

Wei Li

Sundeep Tirumalareddy

Ian C McGraw

Yun-hsuan Sung

Benjamin N Lee

Li Wei

Zhengdong Zhang

Aakanksha Chowdhery

Chenxi Liu

Gregory R Bullock

Slav Petrov

Qiao Liang

Shuo-Yiin Chang

Paul Roland Lambert

Yanping Huang

Eric Lehman

Jacqueline Amy Tsay

Matthew Vincent Dierker

Yuanzhong Xu

Norman Casagrande

Mirkó Visontai

Nan Du

A method for training a non-autoregressive TTS model includes receiving training data that includes a reference audio signal and a corresponding input text sequence. The method also includes encoding the reference audio signal into a variational embedding that disentangles the style/prosody information from the reference audio signal and encoding the input text sequence into an encoded text sequence. The method also includes predicting a phoneme duration for each phoneme in the input text sequence and determining a phoneme duration loss based on the predicted phoneme durations and a reference phoneme duration. The method also includes generating one or more predicted mel-frequency spectrogram sequences for the input text sequence and determining a final spectrogram loss based on the predicted mel-frequency spectrogram sequences and a reference mel-frequency spectrogram sequence. The method also includes training the TTS model based on the final spectrogram loss and the corresponding phoneme duration loss.

Parallel tacotron non-autoregressive and controllable TTS

Systems and methods of the present disclosure are directed to a computing system, including one or more processors and a machine-learned multi-mode speech recognition model configured to operate in a streaming recognition mode or a contextual recognition mode. The computing system can perform operations including obtaining speech data and a ground truth label and processing the speech data using the contextual recognition mode to obtain contextual prediction data. The operations can include evaluating a difference between the contextual prediction data and the ground truth label and processing the speech data using the streaming recognition mode to obtain streaming prediction data. The operations can include evaluating a difference between the streaming prediction data and the ground truth label and the contextual and streaming prediction data. The operations can include adjusting parameters of the speech recognition model.

Systems and methods for training dual-mode machine-learned speech recognition models

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for machine translation using neural networks. In some implementations, a text in one language is translated into a second language using a neural network model. The model can include an encoder neural network comprising a plurality of bidirectional recurrent neural network layers. The encoding vectors are processed using a multi-headed attention module configured to generate multiple attention context vectors for each encoding vector. A decoder neural network generates a sequence of decoder output vectors using the attention context vectors. The decoder output vectors can represent distributions over various language elements of the second language, allowing a translation of the text into the second language to be determined based on the sequence of decoder output vectors.

Machine translation using neural network models

Two-pass automatic speech recognition (ASR) models can be used to perform streaming on-device ASR to generate a text representation of an utterance captured in audio data. Various implementations include a first-pass portion of the ASR model used to generate streaming candidate recognition(s) of an utterance captured in audio data. For example, the first-pass portion can include a recurrent neural network transformer (RNN-T) decoder. Various implementations include a second-pass portion of the ASR model used to revise the streaming candidate recognition(s) of the utterance and generate a text representation of the utterance. For example, the second-pass portion can include a listen attend spell (LAS) decoder. Various implementations include a shared encoder shared between the RNN-T decoder and the LAS decoder.

Two-pass end to end speech recognition

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating waveforms conditioned on phoneme sequences. In one aspect, a method comprises: obtaining a phoneme sequence; processing the phoneme sequence using an encoder neural network to generate a hidden representation of the phoneme sequence; generating, from the hidden representation, a conditioning input; initializing a current waveform output; and generating a final waveform output that defines an utterance of the phoneme sequence by a speaker by updating the current waveform output at each of a plurality of iterations, wherein each iteration corresponds to a respective noise level, and wherein the updating comprises, at each iteration: processing (i) the current waveform output and (ii) the conditioning input using a noise estimation neural network to generate a noise output; and updating the current waveform output using the noise output and the noise level for the iteration.

End-to-end speech waveform generation through data density gradient estimation

Systems and methods can utilize a conformer model to process a data set for various data processing tasks, including, but not limited to, speech recognition, sound separation, protein synthesis determination, video or other image set analysis, and natural language processing. The conformer model can use feed-forward blocks, a self-attention block, and a convolution block to process data to learn global interactions and relative-offset-based local correlations of the input data.

Convolution-augmented transformer models

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for training a neural network to perform any one or more of a variety of machine learning tasks. For example, the neural network can be configured as a generative neural network, e.g., an autoregressive generative neural network.

Training of large neural networks

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training neural networks through contrastive learning. In particular, the contrastive learning is modified to use a relative margin to adjust a training pair's contribution to optimization.

Relative margin for contrastive learning

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 for training a non-autoregressive TTS model includes obtaining a sequence representation of an encoded text sequence concatenated with a variational embedding. The method also includes using a duration model network to predict a phoneme duration for each phoneme represented by the encoded text sequence. Based on the predicted phoneme durations, the method also includes learning an interval representation and an auxiliary attention context representation. The method also includes upsampling, using the interval representation and the auxiliary attention context representation, the sequence representation into an upsampled output specifying a number of frames. The method also includes generating, based on the upsampled output, one or more predicted mel-frequency spectrogram sequences for the encoded text sequence. The method also includes determining a final spectrogram loss based on the predicted mel-frequency spectrogram sequences and a reference mel-frequency spectrogram sequence and training the TTS model based on the final spectrogram loss.

Unsupervised parallel tacotron non-autoregressive and controllable text-to-speech

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating speech from text. One of the systems includes one or more computers and one or more storage devices storing instructions that when executed by one or more computers cause the one or more computers to implement: a sequence-to-sequence recurrent neural network configured to: receive a sequence of characters in a particular natural language, and process the sequence of characters to generate a spectrogram of a verbal utterance of the sequence of characters in the particular natural language; and a subsystem configured to: receive the sequence of characters in the particular natural language, and provide the sequence of characters as input to the sequence-to-sequence recurrent neural network to obtain as output the spectrogram of the verbal utterance of the sequence of characters in the particular natural language.

End-to-end text-to-speech conversion

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for speech synthesis. The methods, systems, and apparatus include actions of obtaining an audio representation of speech of a target speaker, obtaining input text for which speech is to be synthesized in a voice of the target speaker, generating a speaker vector by providing the audio representation to a speaker encoder engine that is trained to distinguish speakers from one another, generating an audio representation of the input text spoken in the voice of the target speaker by providing the input text and the speaker vector to a spectrogram generation engine that is trained using voices of reference speakers to generate audio representations, and providing the audio representation of the input text spoken in the voice of the target speaker for output.

Synthesis of speech from text in a voice of a target speaker using neural networks

Methods, systems, and computer program products for generating, from an input character sequence, an output sequence of audio data representing the input character sequence. The output sequence of audio data includes a respective audio output sample for each of a number of time steps. One example method includes, for each of the time steps: generating a mel-frequency spectrogram for the time step by processing a representation of a respective portion of the input character sequence using a decoder neural network; generating a probability distribution over a plurality of possible audio output samples for the time step by processing the mel-frequency spectrogram for the time step using a vocoder neural network; and selecting the audio output sample for the time step from the possible audio output samples in accordance with the probability distribution.

Synthesizing speech from text using neural networks

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for distributing machine learning workloads, e.g., computations for training a neural network or computing an inference using a neural network, across multiple hardware accelerators. One of the systems comprises a plurality of accelerator islands, each hardware accelerator island comprising a respective plurality of hardware devices that include a plurality of hardware accelerators and a corresponding host for each of the plurality of hardware accelerators; and a respective scheduler for each of the accelerator islands that is configured to schedule workloads across the plurality of accelerators and corresponding hosts in the accelerator island, wherein the system is configured to: receive data representing a machine learning workload; and assign a respective portion of the machine learning workload to each of the plurality of accelerator islands for scheduling by the respective scheduler for the accelerator island.

Asynchronous distributed data flow for machine learning workloads

A method for performing speech recognition using sequence-to-sequence models includes receiving audio data for an utterance and providing features indicative of acoustic characteristics of the utterance as input to an encoder. The method also includes processing an output of the encoder using an attender to generate a context vector, generating speech recognition scores using the context vector and a decoder trained using a training process, and generating a transcription for the utterance using word elements selected based on the speech recognition scores. The transcription is provided as an output of the ASR system.

Speech recognition with sequence-to-sequence models

Growing community of inventors

Fremont, CA, United States of America

Yonghui Wu