| 语种辨识的多特征信息应用 |
| |
| 引用本文： | 曾秀花,杨鉴,徐永华.语种辨识的多特征信息应用[J].计算机工程与应用,2010,46(25):146-148. |
| |
| 作者姓名： | 曾秀花 杨鉴 徐永华 |
| |
| 作者单位： | 1.曲靖师范学院 物理与电子工程学院，云南 曲靖 655011 2.云南大学 信息学院，昆明 650091 |
| |
| 摘 要： | 提取动态的高层语言学特征建立了改进的语种相关的、联合的GMM-LM语种辨识方案。该方案减小了不同语种的高斯混合模型和语言模型之间的相关性，也降低了训练的复杂度。还提出了基于特征提取层和判决层融合技术的语种辨识系统。该系统利用了不同类型的特征对区分不同语种的贡献来增加不同语种语料之间的差异，并使相同语种的语料之间的差异减小。实验表明，设计的语种辨识系统具有较好的扩展性；基于特征提取层和判决层的融合系统能够有效地提高系统识别率。
|
| 关 键 词： | 语种识别 高斯混合模型 特征融合 |
| 收稿时间： | 2009-12-9 |
| 修稿时间： | 2010-4-13
|
Application of multiple speech features to language identification |
| |
| ZENG Xiu-hua,YANG Jian,XU Yong-hua.Application of multiple speech features to language identification[J].Computer Engineering and Applications,2010,46(25):146-148. |
| |
| Authors: | ZENG Xiu-hua YANG Jian XU Yong-hua |
| |
| Affiliation: | 1.School of Physics & Electronic Engineering，Qujing Normal University，Qujing，Yunnan 655011，China 2.School of Information Science and Engineering，Yunnan University，Kunming 650091，China |
| |
| Abstract: | Dynamic high-level language cues are extracted to build the language-dependent and combined GMM-LM language identification system which decreases the correlation between Gaussian Mixture Model and Language Model.It also reduces the training complexity.The feature-level and decision-level multiple features fusion is proposed to the GMM-LM system which distinguishes multiple languages using different feature cues.This system can maximize the speech diversity of different languages and minimize the speech diversity of same languages.It concludes that this modified Gaussian Mixture Model recognizer followed by Language Model has better system expansibility; meanwhile the fusion systems based on feature-level and decision-level can get better performance. |
| |
| Keywords: | language recognition Gaussian mixture model feature fusion |
| 本文献已被 维普 万方数据 等数据库收录！ |
| 点击此处可从《计算机工程与应用》浏览原始摘要信息 |
|
点击此处可从《计算机工程与应用》下载全文 |
