The Transformation of F_0 contour from source speaker to target speaker is a key issue of the high quality voice trans- formation.
基音轨迹F_0～t转换是实现高质量源-目标说话人声音转换的重要组成部分。
One of the easiest ways to test the program is by splicing together various 10- to 30-second intervals of speech from your single-speaker source files.
测试程序的最简单方法之一是从各个单一演讲者的源文件中抽取10 秒到30 秒间隔的声音片段，然后拼接在一起。
But if the goal is to reproduce sounds accurately, then a speaker must be either an acoustic point source (monopole, omni-directional) or an acoustically small bi-directional source (dipole).
但是，如果目标是精确的重放声音的话，一个扬声器必须是声学上的点声源(单极点，全向响应)，或者是一个声学上小的双向源(偶极)。
Interpreting activities involve three parties, I? E. the source language speaker, the interpreter and the target language audience.
口译活动涉及源语发言人、译员和译语听众三方。
Voice conversion is a method which transforms the source speech to converted speech with character information of target speaker.
声音转换是将源说话人的声音转化成具有目标说话人特征信息的声音的方法。
Make waves with a dripping faucet, audio speaker, or laser! Add a second source or a pair of slits to create an interference pattern.
通过滴水的龙头，扬声器或镭射来产生波！增加第二来源或者两条缝，可以产生干涉形式。
For multiple channel audio, use the spread option on the audio Source to spread and split out the discrete channels in speaker space.
对于多声道音频，使用的音频源(Audio Source)中的扩散(spread)选项来扩散并在扬声器空间分离出离散通道。
To obtain clearer desired signal under multi-speaker environment, the time-frequency masking effect is used in post-processing of speech enhancement using blind source separation.
本文将时频掩蔽效应引入盲源分离语音增强系统中，给出一种基于时频掩蔽效应和盲源分离的语音增强方法，并将其与期望最大化方法进行了性能比较。
A new sound source localization and tracking method based on microphone array is presented for speaker tracking problem in noisy and reverberant environments.
一个新的声源定位与跟踪方法，提出了基于麦克风阵列扬声器跟踪问题在嘈杂和混响环境。
A new sound source localization and tracking method based on microphone array is presented for speaker tracking problem in noisy and reverberant environments.
一个新的声源定位与跟踪方法，提出了基于麦克风阵列扬声器跟踪问题在嘈杂和混响环境。
应用推荐