基于感知的右视点质量可分级编码算法

为了减少立体视频传输如此庞大的数据量,根据人眼的立体掩蔽效应,提出了基于感知的右视点质量可分级编码算法．该算法将宏块分为3类：平坦块、纹理块和边缘块,通过不同的量化步长使得右视点图像块的质量可分级．实验证明根据左视点的图像质量,右视点在人眼感知范围内码率减少25％～37％．该算法能在人眼主观察觉范围内很好地提高码率,进一步消除人眼立体视觉冗余．     

南下访“淇淇”

介绍甸独遥长江淡水白暨豚“淇淇”及其同种的发声和听觉的特性,包括它们所发射超声信号（的答声）和可听声信号（哨叫声）以及它们的听觉区域,也介绍了白暨豚当前频临灭绝的处境 。     

一种基于视觉掩盖模型的数字水印技术

通过分析人的视觉特性提出一种视觉掩盖模型,得到该模型水印嵌入算法和水印检测算法。在DCT系统低频域上给典型的“Lena”图像嵌入水印后,进行压缩、加高斯噪声污染、滤波等试验,测试水印的稳健性。结果发现：水印输出响应非常高,有很好的鲁棒性,嵌入水印后的图像经一定的压缩、噪声污染、滤波等处理后仍可被检测到。     

利用掩蔽系数同时测定中草药中的钼和铜(PF-PAN-TritonX-100体系)

本文研究了在混合显色剂（ＰＦ＋ＰＡＮ）的胶束体系中，利用掩蔽系数同时测定中草药中钼和铜的光度分析法．实验表明：在混合体系中，钼量在０～１０μｇ，铜量在０～５０μｇ的范围内符合比耳定律，ε′Ｍｏ＝９２×１０４，ε′Ｃｕ＝１５×１０４．该法用于中草药样品的分析，结果令人满意，回收率可达９２％～１１１％．     

Voice Analysis of Postlingually Deaf Adults Pre- and Postcochlear Implantation

Objectives

To ascertain whether cochlear implantation (CI), without specific vocal rehabilitation, is associated with changes in perceptual and acoustic vocal parameters in adults with severe to profound postlingual deafness.

Hypothesis

Merely restoring auditory feedback could allow the individual to make necessary adjustments in vocal pattern.

Study Design

Prospective and longitudinal.

Methods

The experimental group composed of 40 postlingually deaf adults (20 males and 20 females) with no previous laryngeal or voice disorders. Participants’ voices were recorded before CI and 6–9 months after CI. To check for chance modifications between two evaluations, a control group of 12 postlingually deaf adults, six male and six female, without CI was also evaluated. All sessions composed of the recording of read sentences from Consensus Auditory-Perceptual Evaluation of Voice and sustained vowel /a/. Auditory and acoustic analyses were then conducted.

Results

We found a statistically significant reduction in overall severity, strain, loudness, and instability in auditory analysis. In vocal acoustic analysis, we found statistically significant reduction fundamental frequency (F0) values (in male participants) and F0 variability (in both genders). The control group showed no statistically significant changes in most vocal parameters assessed, apart from pitch and F0 (in female participants only). On comparing the interval of variation of results between the experimental and control groups, we found no statistically significant difference in vocal parameters between CI recipients and nonrecipients, with the exception of F0 variability in male participants.

Conclusions

The patients in our sample showed changes in overall severity, strain, loudness, and instability values, and reductions in F0 and its variability. On comparing the variation of results between the groups, we were able to prove in our study that implant recipients postlingually deaf adults (experimental group), without specific vocal rehabilitation, differed from nonrecipients (control group) in loudness and F0 variability sustained vowel /a/ in male participants.     

一种抗高阶旁路攻击的LED密码算法

LED密码算法是2011年提出的超轻量级密码算法,主要是为资源受限下物联网加密应用研发的.轻量级密码算法结构相对简单,更容易被旁路攻击成功.随机掩码是一种有效抗旁路攻击的方法,在深入LED密码算法结构研究的基础上,提出一种全随机掩码的LED密码算法CMLED.论述了CMLED算法的设计方法,从形式化方面给出了抗高阶旁路攻击选择掩码的原则.同时,对全随机掩码的CMLED与原始算法进行了硬件资源占用与加密效率对比,实验表明CMLED仍然可以高效地在智能卡上实现.     

基于听觉特性的电子耳蜗语音编码策略

为了改善电子耳蜗语音编码策略的性能,提出了一种基于听觉特性的编码策略.策略中的带通滤波及包络检测阶段采用Gammatone滤波器模型和Meddis内毛细胞模型;在通道选择阶段引入心理声学的掩蔽效应模型,从M个通道中选出N个通道,在相应的通道产生脉冲刺激信号.用MATLAB对提出的策略进行仿真研究,实现了22选8的编码策略.结果表明,所提编码策略在稳定性和鲁棒性方面优于CIS和ACE编码策略.     

言语知觉中的感觉运动整合机制

 大脑如何感知和理解言语是心理学和脑科学研究的重要问题之一。近年来,言语产出相关的皮层言语运动系统（SMS）通常被认为以代偿调节方式参与有噪音干扰和/或信息缺失情况下的言语知觉中。本研究采用功能性核磁共振成像（fMRI）技术考查在噪音干扰下的音节辨识任务中的感觉运动整合机制。研究1发现,SMS的参与程度随语音辨识难度增加而增加、在噪音干扰下比听皮层具有更高的音位特异性编码和表征能力,并且在难度适中时对听皮层编码代偿作用最大。研究2发现,听力正常的老年人在噪音干扰下存在音节辨识困难,且比年轻人更加依赖SMS的活动增加以及较少受到老化和噪音影响的言语运动表征以代偿听觉加工能力的下降。本研究不仅加深了对言语知觉中跨通道感觉运动信息整合机制及其毕生发展机制的认识,还为帮助听力受损或正常的老年人设计和开展言语交流康复训练方案提供了新的视角和启发。     

基于雄蝉生物发声机理的蝉鸣声数字化建模研究

目前临床治疗耳鸣的常用方法是耳鸣掩蔽疗法.国内外传统的耳鸣治疗仪普遍存在耳鸣匹配声源不足以及匹配效果差的缺陷.针对该问题,本文基于雄蝉的生物发声机理,通过分析研究其发声过程,从信号处理的角度提出一种模拟雄蝉发声的模型,用以合成最常见的耳鸣声源——蝉鸣声.实验表明,该模型取得了较好的效果,丰富了蝉鸣声匹配声源,具有一定的医用参考价值.     

Significance of Auditory and Kinesthetic Feedback to Singers'' Pitch Control

An accurate control of fundamental frequency (F0) is required from singers. This control relies on auditory and kinesthetic feedback. However, a loud accompaniment may mask the auditory feedback, leaving the singers to rely on kinesthetic feedback. The object of the present study was to estimate the significance of auditory and kinesthetic feedback to pitch control in 28 students beginning a professional solo singing education. The singers sang an ascending and descending triad pattern covering their entire pitch range with and without masking noise in legato and staccato and in a slow and a fast tempo. F0 was measured by means of a computer program. The interval sizes between adjacent tones were determined and their departures from equally tempered tuning were calculated. The deviations from this tuning were used as a measure of the accuracy of intonation. Statistical analysis showed a significant effect of masking that amounted to a mean impairment of pitch accuracy by 14 cent across all subjects. Furthermore, significant effects were found of tempo as well as of the staccato/legato conditions. The results indicate that auditory feedback contributes significantly to singers' control of pitch.