检索规则说明:AND代表“并且”;OR代表“或者”;NOT代表“不包含”;(注意必须大写,运算符两边需空一格)
检 索 范 例 :范例一: (K=图书馆学 OR K=情报学) AND A=范并思 范例二:J=计算机应用与软件 AND (U=C++ OR U=Basic) NOT M=Visual
作 者:汪丹丹[1] WANG Dan-dan(Department of Information Engineering,Anhui Vocational College of City Management,Hefei 230011,China)
机构地区:[1]安徽城市管理职业学院信息工程系,合肥230011
出 处:《西安文理学院学报(自然科学版)》2018年第6期64-67,共4页Journal of Xi’an University(Natural Science Edition)
摘 要:对于各类TTS(Text to Speech)系统而言,能否准确地预测韵律短语边界对TTS系统的效果有着关键性的影响.目前常使用决策树来做韵律短语边界预测,但这种方法受到了训练数据的均衡性以及决策算法本身无法达到全局最优的制约.为了改善预测效果,在传统的决策树方法之上,将决策树使用的聚类属性与模糊决策相结合,提出通过多属性模糊决策方法来预测英文韵律短语边界.实验表明,使用这种方法后,效果比基于决策树的预测方法的效果有较大提升,F-Score由64. 4%提升到69. 3%,不可接受率也从28. 6%降低到21. 4%.For all kinds of TTS(Text to Speech) systems,whether the prosodic phrase boundaries can be accurately predicted has a crucial impact on the TTS system. At present,decision tree is often used to predict the boundaries of prosodic phrases,but this method is restricted by the balance of training data and the decision algorithm itself can not achieve global optimum.Therefore,in order to improve the prediction effect,a multi-attribute fuzzy decision-making method to predict the boundaries of English prosodic phrases based on the traditional decision tree method is proposed in this paper,which combines the clustering attributes used in the decision tree with fuzzy decision-making. Experiments show that the effect of this method is better than that of the decision tree-based prediction method. The F-Score is increased from 64. 4%to 69. 3%,and the unacceptable rate is reduced from 28. 6% to 21. 4%.
关 键 词:英文语音合成 韵律短语边界 决策树 多属性模糊决策
分 类 号:TP391.4[自动化与计算机技术—计算机应用技术]
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在链接到云南高校图书馆文献保障联盟下载...
云南高校图书馆联盟文献共享服务平台 版权所有©
您的IP:216.73.216.124