Google Inc.

Wolfgang Macherey

Franz Josef Och

Shankar Kumar

Yonghui Wu

Zhifeng Chen

Melvin Jose Johnson Premkumar

Michael Schuster

George Foster

Niki J Parmar

Llion Owen Jones

Ankur Bapna

Orhan Firat

Macduff Richard Hughes

Xu Chen

Christopher James Dyer

Peng Xu

Ron J Weiss

Fadi Biadsy

Awaneesh Verma

Joshua James Estelle

Ye Jia

Roy W Tromble

Junyoung Chung

Colin Andrew Cherry

Naveen Arivazhagan

Te I

Pallavi N Baljekar

Melvin Johnson

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

The present disclosure provides systems and methods that train and use machine-learned models such as, for example, sequence-to-sequence models, to perform direct and text-free speech-to-speech translation. In particular, aspects of the present disclosure provide an attention-based sequence-to-sequence neural network which can directly translate speech from one language into speech in another language, without relying on an intermediate text representation.

Direct speech-to-speech translation via machine learning

Methods, systems, and apparatus, including computer programs encoded on computer-storage media, for re-translation for simultaneous, spoken-language machine translation. In some implementations, a stream of audio data comprising speech in a first language is received. A transcription for the speech in the stream of audio data is generated using an automated speech recognizer through a series of updates. A translation of the transcription into a second language is generated using a machine translation module. The translation is generated with translation iterations that translate increasing amounts of the transcription, including re-translating previously portions of the transcription. A series of translation updates are provided to a client device based on the translation iterations.

Re-translation for simultaneous, spoken-language machine translation

Methods, systems, and apparatus, including computer programs encoded on computer storage media for performing multi-task learning. In one method a system obtains a respective set of training data for each of multiple machine learning tasks. For each of the machine learning tasks, the system configures a respective teacher machine learning model to perform the machine learning task by training the teacher machine learning model on the training data. The system trains a single student machine learning model to perform the multiple machine learning tasks using (i) the configured teacher machine learning models, and (ii) the obtained training data.

Multi-task learning using knowledge distillation

Systems, methods, and apparatuses including computer program products for machine translation. A method is provided that includes generating a plurality of machine translation systems using a single machine translation engine, and generating a consensus translation from a plurality of candidate translations for a source sentence, where each candidate translation of the plurality of candidate translations is an output of a respective machine translation system of the plurality of machine translation systems.

Consensus translations from multiple machine translation systems

Methods, systems, and apparatus, including computer program products, for language translation are disclosed. In one implementation, a method is provided. The method includes accessing a hypothesis space, where the hypothesis space represents a plurality of candidate translations; performing decoding on the hypothesis space to obtain a translation hypothesis that minimizes an expected error in classification calculated relative to an evidence space; and providing the obtained translation hypothesis for use by a user as a suggested translation in a target translation.

Optimizing parameters for machine translation

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for presenting alternative translations. In one aspect, a method includes receiving source language text; receiving translated text corresponding to the source language text from a machine translation system; receiving segmentation data for the translated text, wherein the segmentation data includes a first segmentation of the translated text, the first segmentation dividing the translated text into two or more segments; receiving one or more alternative translations for each of the two or more segments; presenting the source text and the translated text to a user in a user interface; and in response to a user selection of a first portion of the translated text, displaying, in the user interface, one or more alternative translations for a first segment to which the first portion of translated text corresponds according to the first segmentation.

Providing alternative translations

Systems, methods, and computer program products are provided for statistical machine translation. In some implementations a method is provided. The method includes receiving multi-lingual parallel text associating a source language, a target language, and one or more bridge languages, determining an alignment between the source language and the target language using a first bridge language that is distinct from the source language and the target language, and using the determined alignment to generate a candidate translation of an input text in the source language to the target language.

Word alignment with bridge languages

Methods, systems, and apparatus, including computer program products, for language translation are disclosed. In one aspect, a method includes accessing a translation hypergraph that represents a plurality of candidate translations, the translation hypergraph including a plurality of paths including nodes connected by edges; calculating first posterior probabilities for each edge in the translation hypergraph; calculating second posterior probabilities for each n-gram represented in the translation hypergraph based on the first posterior probabilities; and performing decoding on the translation hypergraph using the second posterior probabilities to convert a sample text from a first language to a second language.

Growing community of inventors

Mountain View, CA, United States of America

Wolfgang Macherey