Speech synthesis device and speech synthesis method

Abstract

A speech synthesis device includes: a mouth-opening-degree generation unit which generates, for each of phonemes generated from input text, a mouth-opening-degree corresponding to oral-cavity volume, using information generated from the text and indicating the type and position of the phoneme within the text, such that the generated mouth-opening-degree is larger for a phoneme at the beginning of a sentence in the text than for a phoneme at the end of the sentence; a segment selection unit which selects, for each of the generated phonemes, segment information corresponding to the phoneme from among pieces of segment information stored in a segment storage unit and including phoneme type, mouth-opening-degree, and speech segment data, based on the type of the phoneme and the generated mouth-opening-degree; and a synthesis unit which generates synthetic speech of the text, using the selected pieces of segment information and pieces of prosody information generated from the text.

Description

technical field [0001] The present invention relates to a speech synthesis device capable of generating natural synthesized speech and a method thereof. Background technique [0002] Recently, thanks to the development of speech synthesis technology, it is possible to generate very high-quality synthesized speech. As a speech synthesis device with a high sense of authenticity, there is a waveform-connected speech synthesis device in which speech waveforms are selected from a large-scale segment storage unit and concatenated (see, for example, Patent Document 1). Figure 17 It is a typical configuration diagram of a wave connection type speech synthesis device. [0003] Figure 17 The illustrated speech synthesis device includes a speech analysis unit 501 , a prosody generation unit 502 , a speech segment DB (database) 503 , a segment selection unit 504 , and a waveform connection unit 505 . [0004] The language analysis unit 501 linguistically analyzes the input text, and ...

AI Technical Summary

Problems solved by technology

[0015] However, it is difficult to determine the sound quality of synthesized speech only by the above-mentioned phoneme environment and prosodic information

[0016] The inventors of the present application have found the problem that if the time change of the utterance form is different from the time change of the input speech, the naturalness of the change of the utterance form in the synthesized speech cannot be maintained. As a result, the naturalness of the synthesized speech is greatly reduced. deterioration

Images