| 摘要 |
PROBLEM TO BE SOLVED: To provide a smooth utterance transmission technique by segmentation in utterance units in an environment wherein network delay is concealed. SOLUTION: Although a sound of general internet telephone becomes disconnected in a case of a large network delay, the sound is transmitted in syllables in this case to transmit the sound without breaking a pronunciation of each word while the sound is possibly broken in syllable units. There is a problem that when a speech of a simple electric signal is sectioned, a sound, such as a double consonant (glottal stop regarded as a phoneme-"ttsu" in Japanese) and a syllabic nasal ("n" in Japanese), which is in one group in terms of conversation, but small as a speech signal is carelessly sectioned, so a pause is dynamically inserted for each double consonant, each syllabic nasal, and each syllable by using a dictionary and an utterance is transmitted while segmented, thereby realizing the speech transmitting technique for generating no unnatural break halfway in a syllable or a scene of a conversation. COPYRIGHT: (C)2006,JPO&NCIPI
|
