Phoneme information synthesis device, voice synthesis device, and phoneme information synthesis method

Abstract

Provided is a phoneme information synthesis device, including: an operation intensity information acquisition unit configured to acquire information indicating an operation intensity; and a phoneme information generation unit configured to output phoneme information for specifying a phoneme of a singing voice to be synthesized based on the information indicating the operation intensity supplied from the operation intensity information acquisition unit.

Description

[0001] This application claims priority from Japanese application JP 2014-211194, the contents of which are incorporated herein by reference. technical field [0002] The present invention relates to speech synthesis technology, in particular to a technology for synthesizing singing speech in real time according to the operation of an operating member. Background technique [0003] In recent years, with the popularization of speech synthesis technology, there has been an increasing demand for realizing "song performance" by mixing and playing a musical sound signal output from an electronic musical instrument such as a synthesizer and a singing speech signal output from a speech synthesis device. Therefore, speech synthesis apparatuses using various speech synthesis techniques have been proposed. [0004] Here, in order to synthesize singing voices having various phonology and pitches, it is necessary to designate the phonology and pitch of the singing voices, which are synt...

AI Technical Summary

Problems solved by technology

[0005] However, in the first technique, a device capable of inputting characters such as a personal computer is required

Therefore, not only the size of the device becomes larger, but also the cost increases accordingly

In addition, it is difficult for foreigners who do not understand Japanese to input Japanese lyrics

In addition, in English, the same character is sometimes pronounced with different phonemes depending on the situation (for example, when to appears next to have, the phoneme of ve becomes f), and when such a word is input, it is difficult to Confirm in advance whether the pronunciation is pronounced with the desired phoneme

[0006] The second technique simply repeats the same speech (such as "ラ" (ra)), and cannot generate expressive lyrics

[0007] The third technology may generate meaningless lyrics that the user does not want

However, in the third technique, voices are reproduced randomly, so the same lyrics cannot be reproduced repeatedly.

[0008] In addition, the first to third techniques cannot arbitrarily determine the phoneme and synthesize it in real time as a singing voice with an arbitrary pitch, so there is a problem that impromptu singing synthesis cannot be performed.

Images