| 基于HMM-ANN的咳嗽音识别 |
| |
| 引用本文： | 石锐,郑晓平,何庆华. 基于HMM-ANN的咳嗽音识别[J]. 世界科技研究与发展, 2012, 34(5): 751-753,783 |
| |
| 作者姓名： | 石锐 郑晓平 何庆华 |
| |
| 作者单位： | 1. 重庆大学计算机学院,重庆,400044
2. 第三军医大学大坪医院野战外科研究所,重庆,400042 |
| |
| 基金项目： | 重庆市科技攻关项目（CSTC,2009AC5023）资助. |
| |
| 摘 要： | 通过将ANN(人工神经网络)改进应用到HMM(隐马尔科夫模型),使用Mel频率倒谱系数(MFCC)+帧能量+MFCC一阶差分,二阶差分的结构提取咳嗽音特征参数,HMM输出的所有状态累积概率作为ANN的输入序列进行非线性映射,进而提取新的信息来提高HMM的识别性能.实验证明,利用HMM-ANN混和模型来处理咳嗽声识别具有更高的识别精度和可靠性.
|
| 关 键 词： | 隐马尔可夫模型 人工神经网络 特征提取 Mel频率倒谱系数 |
Cough Sound Recognition Based on HMM-ANN |
| |
| SHI Rui , ZHENG Xiaoping , HE Qinghua. Cough Sound Recognition Based on HMM-ANN[J]. World Sci-tech R & D, 2012, 34(5): 751-753,783 |
| |
| Authors: | SHI Rui ZHENG Xiaoping HE Qinghua |
| |
| Affiliation: | 1, College of Computer Science, Chongqing university, Chongqing 400044 ; 2. Surgery Institute, the Third Military Medical University, Chongqing 400042 ) |
| |
| Abstract: | By applying the improved ANN （Artificial Neural Network） to HMM （Hidden Markov Model） , the characteristic parameters is ex- tracted from cough with the structure of frame energy ＋ MFCC ＋ its first-order and second-ordersdifference parameters. All of the stgte cumula- tive probability of HMM outputs will server to be the input sequences of ANN for nonlinear mapping. Then the extraction of new＇information is conducted to improve the recognizing, performance. Experimental results show that compared with the traditional single model, HMM-ANN hybrid mode/has a higher recognizing accuracy and reliability in dealing with cough identification. |
| |
| Keywords: | hidden markov model artificial neural network characteristics extraction mel-frequence eeptral coefficients（MFCC） |
| 本文献已被 维普 万方数据 等数据库收录！ |
|
