联合精确比值掩蔽与深度神经网络的单通道语音增强方法

针对目前有监督语音增强忽略了纯净语音、噪声与带噪语音之间的幅度谱相似性对增强效果影响等问题,提出了一种联合精确比值掩蔽(ARM)与深度神经网络(DNN)的语音增强方法。该方法利用纯净语音与带噪语音、噪声与带噪语音的幅度谱归一化互相关系数,设计了一种基于时频域理想比值掩蔽的精确比值掩蔽作为目标掩蔽;然后以纯净语音和噪声幅度谱为训练目标的DNN为基线,通过该DNN的输出来估计目标掩蔽,并对基线DNN和目标掩蔽进行联合优化,增强语音由目标掩蔽从带噪语音中估计得到;此外,考虑到纯净语音与噪声的区分性信息,采用一种区分性训练函数代替均方误差(MSE)函数作为基线DNN的目标函数,以使网络输出更加准确。实验表明,区分性训练函数提升了基线DNN以及整个联合优化网络的增强效果;在匹配噪声和不匹配噪声下,相比于其它常见DNN方法,本文方法取得了更高的平均客观语音质量评估(PESQ)和短时客观可懂度(STOI),增强后的语音保留了更多语音成分,同时对噪声的抑制效果更加明显。     

基于RLS自适应算法的音频数字水印检测技术

实现了一种基于音频信号掩蔽曲线的水印嵌入方法，提出了一种基于RLS（递归最小二乘方）自适应滤波算法的水印检测算法，该算法在检测水印时无需原音频文件，实验结果表明，在不影响听觉效果的前提下，这种水印嵌入和检测方法对MPEG压缩，常规滤波，加噪等处理均具有很好的抗攻击性即鲁棒性。     

Effects of HearFones on speaking and singing voice quality

HearFones (HF) have been designed to enhance auditory feedback during phonation. This study investigated the effects of HF (1) on sound perceivable by the subject, (2) on voice quality in reading and singing, and (3) on voice production in speech and singing at the same pitch and sound level.

Test 1: Text reading was recorded with two identical microphones in the ears of a subject. One ear was covered with HF, and the other was free. Four subjects attended this test. Tests 2 and 3: A reading sample was recorded from 13 subjects and a song from 12 subjects without and with HF on. Test 4: Six females repeated [pa:p:a] in speaking and singing modes without and with HF on same pitch and sound level.

Long-term average spectra were made (Tests 1–3), and formant frequencies, fundamental frequency, and sound level were measured (Tests 2 and 3). Subglottic pressure was estimated from oral pressure in [p], and simultaneously electroglottography (EGG) was registered during voicing on [a:] (Test 4). Voice quality in speech and singing was evaluated by three professional voice trainers (Tests 2–4).

HF seemed to enhance sound perceivable at the whole range studied (0–8 kHz), with the greatest enhancement (up to ca 25 dB) being at 1–3 kHz and at 4–7 kHz. The subjects tended to decrease loudness with HF (when sound level was not being monitored). In more than half of the cases, voice quality was evaluated “less strained” and “better controlled” with HF. When pitch and loudness were constant, no clear differences were heard but closed quotient of the EGG signal was higher and the signal more skewed, suggesting a better glottal closure and/or diminished activity of the thyroarytenoid muscle.     

耳蜗电图与脑干电反应的综合描记方法

本文同时引导三种听诱发电反应——耳蝸电图(EcochG)、脑干电反应(ABR)和采用以鼓泡后壁为记录电极、头顶部为参考电极的新引导方法引导的EcochG和ABR综合波——E-A综合波。共用12只豚鼠测了16只耳,观察到三种波群相应各波的潜伏期在同样声刺激强度下的潜伏期及随声刺激强度减弱时潜伏期延长的数值和规律相同,说明三者相应各波的起源是同源性的。但E-A综合波具有明显的CM和其后的5～6个波,其中1波最大,即兼有EcochG和ABR两者波形的特点。而且E-A综合波各波波幅也比EcochG和ABR各波波幅大,易于测量。由于E-A综合波具有EcochG和ABR两者的特点,可引导出从耳蜗感受器直到丘脑水平听觉神经系统的电位,所以代表范围广,应在实验研究和临床应用中比EcochG和ABR更有意义,可以代替EcochG和ABR两者的新的引导方法。     

硫代苹果酸作为络合滴定掩蔽剂的研究

本系统地研究了硫代苹果在络合滴定中的掩蔽性能，发现它在ｐＨ１．５～１０可作为Ｓｎ＾４＋、Ｓｂ＾３＋、Ｔｂ＾３＋、Ｂｉ＾３＋、Ｃｕ＾２＋、Ｈｇ＾２＋和Ａｇ＾＋等的掩蔽剂（在ｐＨ２～３．５还可掩蔽Ｆｅ＾３＋）。掩蔽上述离子后，可在ｐＨ２．５～３用Ｔｈ＾４＋回滴ＥＤＴＡ，在ｐＨ５～６用ＥＤＴＡ滴定Ｚｎ＾２＋或Ｙｂ＾３＋等，或在ｐＨ１０用ＥＤＴＡ滴定Ｍｇ＾２＋、Ｃａ＾２＋、Ｍｇ＾２＋＋Ｃａ＾２＋和Ｍｎ＾     

全相位FRM陷波原理及其DSP Builder实现

为设计出陷波点可精确控制、具有高陷波深度及低硬件复杂度的陷波器,提出了一种结合频率响应屏蔽（FRM）技术与全相位滤波技术的高效陷波器实现结构.在分析双相移组合全相位陷波器陷波深度不足的基础上,指出原型滤波器和屏蔽滤波器内含的单窗全相位卷积窗是陷波深度得以大幅度增大的根本原因,理论推导出陷波器频率响应表达式,证明了在平移参数λ取非1/2整数倍的情况下,提出的陷波结构无需任何补偿和校正措施,其陷波深度高达约-200,dB.借助Altera开发工具DSP Builder对所提出的陷波结构进行建模、仿真,产生QuartusⅡ能够识别的VHDL源程序,利用ModelSim进行RTL级仿真,综合、适配并下载至FPGA芯片.实验结果验证了设计方法的正确性和实用性.     

A canonical model for gradient frequency neural networks

We derive a canonical model for gradient frequency neural networks (GFNNs) capable of processing time-varying external stimuli. First, we employ normal form theory to derive a fully expanded model of neural oscillation. Next, we generalize from the single oscillator model to heterogeneous frequency networks with an external input. Finally, we define the GFNN and illustrate nonlinear time-frequency transformation of a time-varying external stimulus. This model facilitates the study of nonlinear time-frequency transformation, a topic of critical importance in auditory signal processing.     

循环掩码型AES密码芯片的相关性功耗攻击

针对循环掩码型AES密码芯片,提出了一种相关性功耗攻击方法.首先利用方差分析找出明文盲化和S盒计算的时间点,然后通过比较盲化后的字节的汉明权重和实际测量功耗值的相关系数恢复掩码,最后针对S盒输入和输出的汉明距离进行相关能量分析.随机选取公开数据集DPA Contest V4提供的1×104条样本曲线进行实验,结果表明该方法仅需要45条曲线即可正确恢复密钥.     

低码率语音编码中感知加权滤波器的一种新型算法

声学心理模型是对人听觉系统生理结构和人耳主观感知特性的数学抽象模型．这种模型已成功运用于宽带音频编码中．首次提出将声学心理模型原理应用于语音编码中感知加权滤波器的设计．通过对ＩＴＵＧ．７２３．１６．３／５．３ｋｂ／ｓ双速率编码算法中感知加权滤波器的改进,编码器的ＭＯＳ分可以改善０．１～０．３,而且新算法所要求的计算量仅比原算法大０．２６ＭＩＰＳ．基于声学心理模型的感知加权滤波器,与现有各类语音编码器中所用的感知加权滤波器相比,有自适应强、更符合人耳听觉特性、主观处理效果更佳的优点．