CAJ | 학술논문

针对传统语音认证算法不适用于资源受限的移动通信终端语音通信的实时性要求，提出了一种高效的基于线性预测分析的语音认证算法。该算法基于语音信号的线性预测分析的原理对LPC系数参数进行优化，并对构成的LPC系数矩阵进行后处理、非负矩阵分解和量化，形成感知特征序列。实验结果表明：所提算法对于内容保持操作具有良好的鲁棒性，并且运算效率没有降低，可以满足语音通信实时性的要求。
침대전통어음인증산법불괄용우자원수한적이동통신종단어음통신적실시성요구，제출료일충고효적기우선성예측분석적어음인증산법。해산법기우어음신호적선성예측분석적원리대LPC계수삼수진행우화，병대구성적LPC계수구진진행후처리、비부구진분해화양화，형성감지특정서렬。실험결과표명：소제산법대우내용보지조작구유량호적로봉성，병차운산효솔몰유강저，가이만족어음통신실시성적요구。
According to the situation that traditional speech authentication algorithms does not appropriated for present speech communication,a speech authentication algorithm based on linear prediction analysis is proposed,it can satisfy the requirement of the efficiency and the robustness for speech authentication. Firstly,the LPC coefficients are optimized based on the principle of linear prediction analysis for speech signal,and make Non-negative Matrix Factorization (NMF) toward the coefficients. Finally,the formed getting Characteristics sequences is quantified. Experiments show that the proposed algorithm has good robustness for content preserving operations,and it doesn't reduce the efficiency while meeting the robustness,it can satisfy the real-time requirement of speech communication.