基于语音特征的自适应压缩感知.pdfVIP

Vo1.33 No.l 第 33 卷第1 期 南京邮电大学学报{自然科学版) 2013 年 2 月 Journal of Nanjing University of Posts and Telecornmunications{ Natural Science) Feh. 2013 基于语音特征的自适应压缩感知 刘智威，吴晓欢，朱卫平 (南京邮电大学信号处理与传输研究院，江苏南京 21C削3) 摘 要:利用语音在 D口域的稀疏性，提出了一种基于语音分为清音和浊音的特点，自适应分配观测点数的语音重 构方法。首先根据清浊音在整个语音段占有的能量比分配观，测点，然后判断每帧语音性质。如果是清音，则根据 能零比的大小来分配该帧的观测点数;如果是浊音，则根据能量的大小来分配观测点数。实验表明:语音信号是稀 疏的并且可压缩，在同种压缩比下，文中所采用的语音重构算法具有较好的信噪比、误差以及MOS 分。 关键词:压缩感知;语音重构;稀疏性;OMP; 自适应 中固分类号:TN912.3 文献标志码:A 文章编号: 1673-5439{2013)01-(泊39-05 Adaptive Compressed Sensing ßased on Speech Characteristics LIU Zhi-wei ,WU Xiao-hu钮，ZHU Wei-ping ( Institute of Signal proc回sing 皿d Transmission ,Nanjing University of Posts and Telecommunications ， N皿ljing 21【削3 ， China) Abstract: Utilizing the sp缸sity of the speech signals in the DCT domain and considering the composition of voiced and unvoiced speech ,a speech recovery method is proposed wi由 adaptively distributed number of measurement points. It distributes the measurement points according to the energy ratio of voiced speech and unvoiced speech and then judges on 出e prope向 of each speech frame. H the frame is unvoiced speech ,the number of measurement points is calculated according to its zeros and energy rate. H the frame is voiced speech ,the number of measurement points is calculated according to its energy. 币le experiment results demonstrate that the recovered speech signal resulting from the proposed method has good signal to noise ratio ,high MOS score and low error. Key words: compressed sensing; speech recoveη; sparsity; OMP; adaptability 同 吉 o sl 直接对信号进行较少采样得到信号的压缩表示，从