National Institute of Information and Communications Technology, Incorporated

Eiichiro Sumita

Noriyuki Kimura

Chiori Hori

Satoshi Nakamura

Yutaka Ashikari

Akihiro Tamura

Taro Watanabe

Yutaka Kidawara

Masao Uchiyama

Atsushi Fujita

Kenji Imamura

Kehai Chen

Rui Wang

Improvement is made in performance of a trained neural network that uses positional information indicating a position at which each token included in an input sequence is present in the input sequence. An inferencer includes a first generation unit that generates an intermediate sentence representation based on a first sentence representation having information indicating a value of each token included in the input sequence and first positional information indicating a position at which each token is present in the input sequence; a second generation unit that generates second positional information by modifying the first positional information based on the first sentence representation and the intermediate sentence representation, and that generates a hidden state representation based on the second positional information and the intermediate sentence representation; and a third generation unit that generates a second sentence representation based on the intermediate sentence representation and the hidden state representation.

Inferencer, inference method and inference program

Provided are a model training method for neural machine translation that enhances an encoder using a monolingual corpus of a target language and improves the accuracy of the entire translator, and a machine translation system for performing the model training method. The machine translation systemuses a monolingual corpus of the target language to obtain multiple pieces of pseudo source language data, thus allowing for obtaining a large amount of pseudo parallel corpus data having diversity. Further, the machine translation systemuses both the pseudo parallel corpus data having diversity, which has been obtained in large quantities, and the base parallel corpus data in a small quantity but with high accuracy, with the applied learning rates changed accordingly, to perform the learning process (training process) for the machine translation model. This allows the machine translation systemto obtain a learned model (machine translation model) with very high accuracy.

Pseudo parallel translation data generation apparatus, machine translation processing apparatus, and pseudo parallel translation data generation method

In order to solve a problem that the level of precision in inferring a crosslingual topic of documents or words in a non-parallel corpus associated in the document level is not high, a topic inferring apparatus includes: a word distribution information storage unit in which word distribution information is stored in association with each of two or more languages; a document topic distribution generating unit that acquires document topic distribution information of a multilingual document set; a segment topic distribution generating unit that acquires segment topic distribution information of each segment, using the document topic distribution information; and a word topic determining unit that determines, for each word contained in two or more documents contained in the multilingual document set, a topic of each word using the segment topic distribution information. Accordingly, it is possible to improve the level of precision in inferring a topic.

Topic inferring apparatus, topic inferring method, and storage medium

[Object] An object is to provide an apparatus for attaining highly precise word alignment. [Solution] The apparatus includes: selecting means receiving a bilingual sentence pair and a word alignment for the bilingual sentence pair, for successively selecting words fof a sentence in a first language in a prescribed order; and a recurrent neural network (RNN), computing, for all words of the sentence in the first language, a scorerepresenting a probability that a word pair consisting of the word fand a word ealigned with the word fby a word alignment ain a second language of the bilingual sentence pair is a correct word pair, and based on this score, for computing a score of the word alignment a. When computing a score of word pair (f, e), RNNcomputes a scoreof the word pair (f, e) based on all word alignments aselected by the selecting means prior to the word fof the word pair (f, e), of the word alignments a, by means of a recurrent connection

Word alignment score computing apparatus, word alignment apparatus, and computer program

A conventional speech recognition dictionary, translation dictionary and speech synthesis dictionary used in speech translation have inconsistencies. Inconsistency among dictionaries can be solved with a dictionary server including an all-language group dictionary storage unit capable of having stored therein, for two or more languages, two or more all-language term information pieces for speech recognition, translation and speech synthesis, an information-for-speech-recognition sending unit that acquires information for speech recognition including speech recognition information of a term for all languages, and sends the information to one or more speech recognition servers, an information-for-translation sending unit that acquires information for translation including the written form of the term for all languages, and sends the information to one or more translation servers, and an information-for-speech-synthesis sending unit that acquires information for speech synthesis including speech synthesis information of a term for all languages, and sends the information to one or more speech synthesis servers.

Speech translation system, dictionary server, and program

Appropriate processing results or appropriate apparatuses can be selected with a control device that selects the most probable speech recognition result by using speech recognition scores received with speech recognition results from two or more speech recognition apparatuses; sends the selected speech recognition result to two or more translation apparatuses respectively; selects the most probable translation result by using translation scores received with translation results from the two or more translation apparatuses; sends the selected translation result to two or more speech synthesis apparatuses respectively; receives a speech synthesis processing result including a speech synthesis result and a speech synthesis score from each of the two or more speech synthesis apparatuses; selects the most probable speech synthesis result by using the scores; and sends the selected speech synthesis result to a second terminal apparatus.

Speech translation system, control device, and control method

In conventional network-type speech translation systems, devices or models for recognizing or synthesizing speech cannot be changed in accordance with speakers' attributes, and therefore, accuracy is reduced or inappropriate output occurs in each process of speech recognition, translation, and speech synthesis. Accuracy of each processing of speech translation, translation, or speech synthesis is improved and appropriate output is performed in a network-type speech translation system by, based on speaker attributes, appropriately changing the server to perform speech recognition or the speech recognition model, appropriately changing the translation server to perform translation or the translation model, or appropriately changing the speech synthesis server or speech synthesis model.

Growing community of inventors

Koganei, Japan

Eiichiro Sumita