低声语音密码算法
首发时间:2014-10-28
摘要:智能手机获得巨大发展和普及的同时,也存在着巨大的安全隐患。密码作为安全保护的主要手段之一,一直备受重视。由于传统的文本密码具有容易忘记及偷窃的缺点,非文本密码的研究在近年逐渐获得重视。本文的研究内容是利用低声语音这种非文本密码作为手机密码。工作主要包括:(1)说话人自动识别系统(Automatic Speaker Recognition, ASR)的概述以及其包含的各项技术,如基频提取、线性预测倒谱系数(Linear Prediction Cepstrum Coefficient, LPCC)特征提取、动态时间规整(DTW)的原理等;(2)语音库制作;(3)低声语音识别密码算法的研究。实验结果表明,本文提出的基于LPCC-DTW的低声语音密码算法具有一定的识别功能。
For information in English, please click here
Agorithm of Whisper as Password
Abstract:Today, many risks are presenting challenges to security when smart phones have been widely spread out. Password is valued as a method of security protection. Since traditional plain-text password is apt to be forgotten and stolen, the study of non-text password has drawn more attention in recent years. In this paper, algorithm of smart phone password using whispered voice is investigated. The main contributions of this thesis are as follows. (1) An overview of automatic speaker recognition(ASR) and other techniques, such as pitch extraction, LPCC(Linear Prediction Cestrum Coefficient), and DTW(Dynamic Time Warping) is given. (2) Production of the corpus for this thesis is introduced. (3) The algorithm of whisper voice code recognition is proposed. (4) Experiments are conducted and discussions are given. The experiment results indicate that the algorithm of whisper voice code recognition proposed in this paper achieves good recognition performances.
Keywords: Whisper Speaker Recognition LPCC DTW
论文图表:
引用
No.4613104660775141****
同行评议
勘误表
低声语音密码算法
评论
全部评论0/1000