Estimating speaker-specific intonation patterns using the linear alignment model
Estimating speaker-specific intonation patterns using the linear alignment model
复制标题
使用线性对齐模型估计说话者特定的语调模式
DOI：
10.21437/interspeech.2013-99
复制
发表时间：
2013
期刊：
影响因子：
--
通讯作者：
J. V. Santen
中科院分区：
文献类型：
--
作者：
G. Kiss;J. V. Santen
建模特定于说话者的语调在几个领域都很重要，包括说话者识别、验证和使用文本到语音合成的模仿。然而，选择的入境模型和估计其参数的自发语音仍然是一个挑战。我们提出了一种方法来估计特定于说话者的语调参数的一个特定的超命题模型，简化的艾德线性对齐模型[1]，使用鲁棒的每话语和整体统计的自发语音。我们用这种方法比较了自闭症或语言障碍儿童的语调，他们通常有非典型的语音韵律，与正常发育的儿童。我们发现各组之间存在显著差异，这证明了所提出方法的有效性。
Modeling speaker-specific intonation is important in several areas, including speaker identification, verification, and imitation using text-to-speech synthesis. However the choice of the into-nation model and the estimation of its parameters from sponta-neous speech remains a challenge. We propose a way to estimate speaker-specific intonation parameters for a particular su-perpositional model, the Simplified Linear Alignment Model [1], using robust per-utterance and overall statistics of spon-taneous speech. We used this method to compare the intonation of children with autism or language impairment, who often have atypical speech prosody, with that of typically developing children. We found significant differences between the groups, which demonstrates the effectiveness of the proposed method.