资源列表
weitebisuanfa
- 维特比算法 寻找最可能的隐藏状态序列(Finding most probable sequence of hidden states) 对于一个特殊的隐马尔科夫模型(HMM)及一个相应的观察序列，我们常常希望能找到生成此序列最可能的隐藏状态序列。 -Viterbi algorithm to find the most likely hidden state sequence (Finding most probable sequence of hidden states) for a
chaotic_sequence
- CS用于语音信号处理，观测矩阵采用混沌序列-compressed sensing with chaotic sequence
BSStechnology
- 语音信号盲分离技术的研究与应用 硕士论文-BSS technology on voice
yuyin
- 语音自动应答系统，用于应答语音信号。开发环境VC6.0已经测试通过。-Automatic voice response system for the answer voice signals. VC6.0 development environment has been tested.
sphinx_Doc_Chan070111
- 语音识别工具包CMU Sphinx3.0.6的开发文档。-CMU Sphinx3.0.6 speech recognition development kit documentation.
AudioFeatureMFCC
- 能提取音频信号处理中最常用的MFCC特征-Can extract the audio signal processing characteristics of the most commonly used MFCC
008.pdf
- A speech recognition paper.
fellbaum.pdf
- A speech recognition paper.
Soudcode
- 声音验证码，数字0-9，请使用VS2008 SP1或VS2010打开。-Voice verification code, numbers 0-9, use the VS2008 SP1 or VS2010 open.
T-REC-G.722-198811-I!!PDF-E
- G.722协议，可应用与各种高品质语音应用-G.722 protocol, can be applied with a variety of high-quality voice applications
question_bank_ranjith
- speech signal processing question
dtw
- 本系统是一个在多媒体PC 上实现的孤立词识别系统, 它提取语音的线性预测系数作为特征参数, 并采用Itaku ra 失真测度计算帧间距离, 在识别上则使用了动态时轴弯曲(DTW ) 进行时间匹配。本系统对一般的DTW 法作了改进, 即通过放宽端点限制以得到更好的语音匹配, 克服了一般DTW 法要求语音首尾严格对齐而造成的弊病, 降低了语音端点检测的精度要求。-This system is implemented on a multimedia PC isolated word recogniti