| 基于遗传算法的声学模型拓扑结构优化 |
| |
| 引用本文： | 包希日莫,高光来,张 璟. 基于遗传算法的声学模型拓扑结构优化[J]. 计算机工程与应用, 2014, 50(14): 5-8 |
| |
| 作者姓名： | 包希日莫 高光来 张 璟 |
| |
| 作者单位： | 1.内蒙古大学 计算机学院，呼和浩特 0100212.西安理工大学 计算机科学与工程学院，西安 710048 |
| |
| 摘 要： | 针对当前创建语音识别系统时只能采用经验式或启发式方法选择声学模型拓扑结构的情形，提出了一个基于标准遗传算法的声学模型拓扑结构优化算法。与以往的类似应用相比，该算法具备同时优化模型状态数与各状态高斯核数和摒弃高斯核均匀分配的特点。连续数字串TIDigits语料上的以贝叶斯信息准则为目标函数的实验表明，与传统方法创建的基线系统相比，模型拓扑优化的系统能够以较低的复杂度获得较高的识别率，这说明该算法是声学模型拓扑结构优化的有效工具。
|
| 关 键 词： | 隐马尔可夫模型 遗传算法 语音识别 声学模型拓扑结构 贝叶斯信息准则 |
Genetic algorithm based optimization of acoustic model topologies |
| |
| BAO Xirimo,GAO Guanglai,ZHANG Jing. Genetic algorithm based optimization of acoustic model topologies[J]. Computer Engineering and Applications, 2014, 50(14): 5-8 |
| |
| Authors: | BAO Xirimo GAO Guanglai ZHANG Jing |
| |
| Affiliation: | 1.School of Computer Science, Inner Mongolia University, Hohhot 010021, China2.Institute of Computer Science & Engineering, Xi’an University of Technology, Xi’an 710048, China |
| |
| Abstract: | Aiming at the current situation of selecting acoustic model topologies empirically or heuristically, an acoustic model topology optimization algorithm based on standard Genetic Algorithm（GA） is proposed. Compared with the previous similar algorithms, it is an automatic one with the characteristics of optimizing model state number and kernel numbers simultaneously and rejecting uniform allocation of Gaussian kernels. Experiments using Bayesian Information Criterion（BIC） as the objective function on TIDigits corpus show that, the speech recognition system whose model topologies are optimized using the GA based algorithm is able to obtain higher recognition performance with smaller number of parameters compared with the baselines built in the conventional way, which indicates that the GA based algorithm is an effective tool of optimizing acoustic model topologies. |
| |
| Keywords: | Hidden Markov Model（HMM） Genetic Algorithm（GA） speech recognition acoustic model topology Bayesian Information Criterion（BIC） |
| 本文献已被 CNKI 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》下载免费的PDF全文 |
