Nara Institute of Science and Technology

Tomoki Toda

Kiyohiro Shikano

Graham Neubig

Satoshi Nakamura

Yusuke Oda

Mikihiro Nakagiri

Sakriani Sakti

Hideki Kashioka

A translation apparatus that can perform translation based on a correct syntax, a syntax evaluation apparatus employed in the translation apparatus, a syntax evaluation method, and a syntax evaluation program are provided. A syntax evaluation unitincludes a grammatical element guessing partthat guesses a grammatical element that should be connected before and after or after a word string indicated by first fragment data, and a grammatical element complementing partthat complements, based on the guessing result of the grammatical element guessing part, the first fragment data as necessary to generate second fragment data. When the grammatical element guessing partguesses a grammatical element, the grammatical element complementing partcomplements, at a location to which the grammatical element of the word string indicated by the first fragment data should be connected, a tag indicating the grammatical element in order to generate the second fragment data. When the grammatical element guessing partdoes not guess a grammatical element, the grammatical element complementing parthandles the first fragment data as the second fragment data.

Syntax evaluation apparatus, translation apparatus, syntax evaluation method, and syntax evaluation program

[Problems] To convert a signal of non-audible murmur obtained through an in-vivo conduction microphone into a signal of a speech that is recognizable for (hardly misrecognized by) a receiving person with maximum accuracy. [Means for Solving Problems] A speech processing method comprising: a learning step (S) for conducting a learning calculation of a model parameter of a vocal tract feature value conversion model indicating conversion characteristic of acoustic feature value of vocal tract, on the basis of a learning input signal of non-audible murmur recorded by an in-vivo conduction microphone and a learning output signal of audible whisper corresponding to the learning input signal recorded by a prescribed microphone, and then, storing a learned model parameter in a prescribed storing means; and a speech conversion step (S) for converting a non-audible speech signal obtained through an in-vivo conduction microphone into a signal of audible whisper, based on a vocal tract feature value conversion model, with a learned model parameter obtained through the learning step set thereto.

Apparatus and method for producing an audible speech signal from a non-audible speech signal

Growing community of inventors

Nara, Japan

Tomoki Toda