基于ResNet的音频场景声替换造假的检测算法

针对造假成本低、不易察觉的音频场景声替换的造假样本检测问题，提出了基于ResNet的造假样本检测算法。该算法首先提取音频的常数Q频谱系数（CQCC）特征，之后由残差网络（ResNet）结构学习输入的特征，结合网络的多层的残差块以及特征归一化，最后输出分类结果。在TIMIT和Voicebank数据库上，所提算法的检测准确率最高可达100%，错误接收率最低仅为1.37%。在现实场景下检测由多种不同录音设备录制的带有设备本底噪声以及原始场景声音频，该算法的检测准确率最高可达99.27%。实验结果表明，在合适的模型下利用音频的CQCC特征来检测音频的场景替换痕迹是有效的。     

基于小波变换的音频扩频掩密分析算法

提出一种基于小波去噪技术和小波突变点检测技术的音频扩频掩密分析算法。该算法利用小波去噪算法分离含密音频上的噪声信号,循环截取不同长度的噪声信号与剩余的噪声信号进行滑动相关计算,利用小波突变点检测技术检测计算得出的滑动相关值中的突变点,提取这些突变点特征对待检测的音频进行掩密分析。实验结果表明,在PN序列嵌入强度大于0.002时,算法的检测正确率达80%以上。     

基于自适应阈值与基频检测的自发性口语音频分割算法

为了去除自发性口语音频中静音和噪音段的干扰,提高语音识别率和解码识别效率,提出一种音频能量自适应阈值计算方法。针对实时自动口语评测应用,设计了能量阈值自适应系数,该方法将根据能量阈值自适应系数动态地给每个考生的个人单次所有考试音频计算匹配一个能量阈值,以避免阈值选择和硬门限判决造成的误检。在基于自适应能量阀值的音频切分后,加入了基频检测步骤,以判别切分后所得音频段是否为噪声,从而最终分离出纯净的口语音频部分。实验结果表明,该算法能有效准确地切分音频,且鲁棒性较强。     

差分电容电压转换电路噪声性能分析及测试

本文对电容检测式加速度计系统中广泛采用的差分电容电压转换电路建立了电容电压转换电路的等效噪声模型,并对双运放集成电路芯片所构成的差分电容电压转换电路的本底噪声以及仪表放大器输出端的噪声进行了测试,将电容电压转换电路本底噪声中的差模噪声分量和共模噪声分量进行了分离.测试结果表明影响加速度计系统噪声性能的差模噪声分量占电容...     

基于替代数据检测的音频信号非线性分析

针对轻微含噪的音频信号,本文提出了一种基于替代数据检测的非线性分析方法。该方法首先根据音频信号的线性假设生成多组替代数据,然后分别计算原始数据和替代数据的样本峰度,最终根据假设检验方法判断原始音频中是否包含非线性成分。实验结果表明,所提方法能够有效地验证音频信号的非线性特性。相比于传统基于最大Lyapunov指数的非线性分析方法,该方法能更好地区分音频信号和噪声信号。     

通道间串扰测量不确定度评定

针对通道间串扰校准评价离散性较大并且受测量系统本底噪声和偏移影响严重的问题,提出了使用绝对值的平均值计算干扰分量值以降低离散性,通过绝对值幅度合成法剔除本底噪声和偏移的影响,从而获得离散性良好、受本底噪声与直流偏移影响较低的通道间串扰测量结果。并基于该方法与流程,讨论了不确定度的来源,主要包括信号源误差、数据采集系统增益误差、幅度分辨力误差、测量通道直流偏移、干扰本身的测量误差、本底噪声的测量误差和测量重复性等,并建立了不确定度模型。最后结合实例,给出了通道间串扰的不确定度评价结果,验证了所述方法的实用性与可行性。     

漫谈公共影院和家庭影院的还音系统

贵刊今年第2期发表了《家用AV功放真需要百瓦以上的输出功率吗?》一文,由于视听技术的飞跃发展AV设备的应用,已是今非昔比,因此,有些问题,值得“大家共同研讨”。七十年代,上海地区的中型(千座左右)电影院还是单声道的。单声道电影的还音频响为60-8000Hz,载体的本底噪声34dB,动态范围60dB(中频段,下同),音箱用2分频2单元。常用扬声器单元的电声性能见表1。     

基于小波分解和信号相关函数的语音端点检测

为检测和分离含噪语音信号中的信号段和噪声段,提出一种基于小波分解和信号相关函数的检测方法.该方法对含噪信号进行多层小波分解,利用相邻层重构信号间的相似性,通过信号相关计算来检测语音端点.实验表明:该方法能够较准确地在噪声污染的音频中检测出语音端点,其抗噪声干扰能力强于美尔倒谱检测法.     

一种低噪声MEMS加速度计设计与制作

为了提高微加速度计的噪声性能,研究了一种基于绝缘体上硅(SOI)技术的单轴MEMS加速度计的设计和加工方案。该微加速度计采用大面积质量块的电容式检测结构,通过增加检测质量,在保证灵敏度的前提下,有效地降低了微加速度计的机械布朗噪声,增强了信噪比。另外,该微加速度计采用一种基于Al保护层的MEMS SOI工艺技术制造,有利于提高微加速度计的整体精度水平。测试结果表明:微加速度计的本底噪声为20μgn/√Hz,灵敏度为2.5 V/gn。     

一种基于指纹因子的鲁棒音频检索方法

针对基于内容的音频检索中由于噪声造成的查找失败问题,本文提出了一种对噪声鲁棒的基于音频指纹因子的音频特征提取算法和一种半监督的音频字典训练算法,以提高噪声下音频检索的精度。本文方法从Mel谱中提取音频指纹,利用非负矩阵分解算法将指纹分解为对噪声鲁棒的频率因子和时间因子作为特征。同时通过提出的半监督音频字典训练算法进行音频字典训练,本文方法使用音效集计算基本音效的分布空间作为初始字典,在量化数据的同时动态更新字典以实现对数据的准确描述。实验结果表明,在低信噪比条件下本文提出的算法的平均查询精度明显高于其他算法。     

基于噪声一致性的数字语音异源拼接篡改检测算法

异源拼接是一种常见的数字语音篡改行为,其主要借助音频编辑软件将不同场景中录制的语音片段拼接在一起,以达到改变语音语义的目的。考虑到不同场景中所包含的背景噪声特性往往存在差异,提出了一种基于噪声一致性的数字语音异源拼接篡改检测算法。首先,采用时间递归平均（TRA）算法提取待检测语音中所含噪声;然后,通过突变点检测（CPD）算法检测噪声方差是否存在突变来判定待检测语音是否经过篡改,并对篡改位置作出定位。实验仿真结果表明,所提算法能对数字语音中的异源篡改位置进行有效检测。     

基于神经网络的扩展回声隐藏算法

提出了一种神经网络自适应扩展回声隐藏算法。利用PN序列将音频信号的单回声内核进行扩展后作为水印信号，提高了水印算法的安全性。该算法利用了神经网络的非线性映射能力确定扩展回声内核的幅值，从而避免了复杂的心理声学模型的计算过程，实现了水印嵌入的强度的自适应。仿真实验证明了该算法的有效性和可靠性。     

A robust digital audio watermarking based on statistics characteristics

In digital audio watermarking, the watermark's vulnerability to desynchronization attacks has long been a difficult problem. According to the audio statistics characteristics and synchronization code technique, a new robust audio watermarking scheme against desynchronization attacks is proposed in this paper. Firstly, the original digital audio is segmented and then each audio segment is cut into two parts. Secondly, with the spatial watermarking technique, the synchronization code is embedded into the statistics average value of audio samples in the first part. Finally, the second part of audio segment is cut into audio sections, the DWT is performed on the audio sections, and the watermark bit is embedded into the statistics average value of low frequency components. Experimental results show that the proposed scheme is inaudible and robust against common signals processing, including MP3 compression, low-pass filtering, noise addition, and equalization, etc. Moreover, it also survives several desynchronization attacks, such as random cropping, amplitude variation, pitch shifting, time-scale modification, and jittering, etc.     

Echo threshold between passive and electro-acoustic transmission paths in digital hearing protection devices

Electronic hearing protection devices are increasingly used in noisy environments. Theses devices feature a miniaturized external microphone and internal loudspeaker in addition to an analog or digital electronic circuit. They can transmit useful audio signals such as speech and warning signals to the protected ear and can reduce the sound pressure level using dynamic range compression. In the case of a digital electronic circuit, the transmission of audio signals may be noticeably delayed because of the latency introduced by the digital signal processor and by the analog-to-digital and digital-to-analog converters. These delayed audio signals will hence interfere with the audio signals perceived naturally through the passive acoustical path of the device. The proposed study presents an original procedure to evaluate, for two representative passive earplugs, the shortest delay at which human listeners start to perceive two sounds composed of the signal transmitted through the electronic circuit and the passively transmitted signal. This shortest delay is called the echo threshold and represents the delay between the time of perception of one fused sound from two separate sounds. In this study, a transient signal, a clean speech signal, a speech signal corrupted by factory noise, and a speech signal corrupted by babble noise are used to determine the echo thresholds of the two earplugs. Twenty untrained listeners participated in this study, and were asked to determine the echo thresholds using a test software in which attenuated signals are delayed from the original signals in real-time. The findings show that when using hearing devices, the echo threshold depends on four parameters: (a) the attenuation function of the device, (b) the duration of the signal, (c) the level of the background noise and (d) the type of background noise. Defined here as the shortest time delay at which at least 20% of the participants noticed an echo, the echo threshold was found to be 8 ms for a bell signal, 16 ms for clean speech and 22 ms for speech corrupted by babble noise when using a shallow earplug fit. When using a deep fit, the echo threshold was found to be 18 ms for a bell signal and 26 ms for clean speech and 68 ms for speech in factory. No echo threshold could be clearly determined for the speech signal in babble noise with a deep earplug fit.     

A new adaptive digital audio watermarking based on support vector machine

It is a challenging work to design a robust digital audio watermarking scheme against desynchronization attacks. On the basis of support vector machines (SVMs), a new robust digital audio watermarking algorithm against desynchronization attacks is proposed in this paper, and in this the audio statistics characteristics and synchronization code are utilized. Firstly, the optimal embedding positions are located adaptively by using the SVM theory. Secondly, the 16-bit Barker code is chosen as synchronization mark and embedded into the digital audio by modifying the statistics average value of several samples. Finally, the digital watermark are embedded into the statistics average value of low-frequency components in wavelet domain by making full use of auditory masking. Experimental results show that the proposed scheme is inaudible and robust against common signal processing such as MP3 compression, low-pass filtering, noise addition, equalization, etc., and is robust against desynchronization attacks such as random cropping, amplitude variation, pitch shifting, time-scale modification, jittering, etc.     

用于版权保护的鲁棒音频水印

提出一种将数字水印在时间域上嵌入到数字音频信号中的方法，该方法利用混沌序列具有容易生成，对初始条件敏感，以及具有白噪声的统计特生等特点，水印的抽取不需要原始音频信号的知识，实验结果表明，该方法能够保证好的音频信号质量并对大多数普通信号处理具有鲁棒性。     

基于FPGA+DSP架构的数字音频处理技术研究

为了解决航空机载环境下飞行员通话强噪声问题,提出了一种基于FPGA+DSP架构的数字话音处理系统.系统由模拟部分和数字部分组成,模拟部分完成话音信号的匹配、滤波、放大和AD/DA转换;数字部分设计了一种音频处理算法,对话音信号进行活动检测、噪声抑制和话音增强等处理.试验结果表明,该系统能够有效抑制通话噪声、增强话音信号,提高了飞行员通话的可懂度和舒适度.     

数字音频篡改被动检测研究综述

数字音频篡改被动检测是指不依赖任何预先嵌入的信息来鉴别数字音频真伪的技术,其最主要研究内容是判定数字音频的真实性和完整性,在司法取证、新闻公正、知识产权保护等领域有着广泛的应用前景。目前领域内相关综述主要从数字音频主动、被动取证总体框架开展,并未专门针对数字音频篡改被动取证研究进行系统全面总结,且涉及被动取证部分存在时效性不足的问题。据此首先总结了数字音频篡改被动检测的任务模型和取证框架,接着依据篡改手段、检测策略、所使用的统计特征及模型,将目前的数字音频篡改被动检测方法分为四类：基于篡改操作的检测方法、基于数字音频重压缩的检测方法、基于录音设备和音频录制环境的检测方法、基于数字音频信号自身统计特性的检测方法,然后分析了每种方法所采用的典型算法和扩展手段,并对不同检测算法进行性能比较,然后对这四类方法的检测特点和使用范围进行总结。最后综合近年来国内外研究人员的主要成果,总结了数字音频篡改被动检测研究面临的问题和挑战,并对未来的研究进行了展望。     

一种基于倒谱变换的自适应音频水印算法*

结合人类听觉系统,提出了一种基于倒谱变换的自适应音频水印算法,充分利用复倒谱变换的性质,将原始音频信号分成若干帧,每帧实施复倒谱变换后,在对应位置按照一定的方法嵌入水印信号。水印的提取不需要原始音频信号,是一种盲水印算法。实验结果表明,嵌入后的水印不仅具有很好的不可感知性,而且对添加噪声、重新采样、低通滤波和重新量化等攻击也具有很好的鲁棒性。     

A New Adaptive Digital Audio Watermarking Based on Support Vector Regression

On the basis of support vector regression (SVR), a new adaptive blind digital audio watermarking algorithm is proposed. This algorithm embeds the template information and watermark signal into the original audio by adaptive quantization according to the local audio correlation and human auditory masking. The procedure of watermark extraction is as follows. First, the corresponding features of template and watermark are extracted from the watermarked audio. Then, the corresponding feature of template is selected as training sample to train SVR and an SVR model is returned. Finally, the actual outputs are predicted according to the corresponding feature of watermark, and the digital watermark is recovered from the watermarked audio by using the well-trained SVR. Experimental results show that our audio watermarking scheme is not only inaudible, but also robust against various common signal processing (such as noise adding, resampling, requantization, and MP3 compression), and also has high practicability. In addition, the algorithm can extract the watermark without the help of the original digital audio signal, and the performance of it is better than other SVM audio watermarking schemes.