SPEECH SYNTHESIZER, ELECTRONIC WATERMARK INFORMATION DETECTION DEVICE, SPEECH SYNTHESIS METHOD, ELECTRONIC WATERMARK INFORMATION DETECTION METHOD, SPEECH SYNTHESIS PROGRAM, AND ELECTRONIC WATERMARK INFORMATION DETECTION PROGRAM

摘要

Provided are a speech synthesizer, an electronic watermark information detection device, a speech synthesis method, an electronic watermark information detection method, a speech synthesis program, and an electronic watermark information detection program in which an electronic watermark can be inserted without deterioration of sound quality of a synthesized speech. An information processing device of an embodiment includes a sound source generation section, a phase modulation section, and a vocal tract filter section. The sound source generation section generates sound source signals by using basic frequency series of speech and a pulse signal. The phase modulation section modulates phases of pulse signals per pitch mark based on the electronic watermark information, with respect to sound source signals generated by the sound source generation section. The voice tract filter section generates speech signals by using spectrum parameter series, with respect to the sound source signals with the phases of pulse signals modulated by the phase modulation section.