Speaker feature extraction based on human auditory model and speaker identification under noisy background 基于聽覺模型的話者特征參數(shù)提取及其在噪聲背景下的話者辨識
Main ideas as follows : research of audio digital watermark algorithm based on fast fourier transforms and psychoacoustic auditory model 主要工作有:基于快速傅立葉變換與聲學(xué)模型的音頻數(shù)字水印算法研究。
The result of experiments show that resynthesized speech signals form the its correlogram by auditory model inversion is nature and robust in noisy environment 實(shí)驗(yàn)結(jié)果表明,我們通過聽覺模型反演從信號的自相關(guān)圖譜中恢復(fù)出的語音信號,具有較好的自然度和良好的噪聲魯棒性。
With the auditory model as the front - end to extract the correlogram of signals . following , this paper present the implementation of suditory model inversion procedure by resynthesizing original signal from the correlogra - m 接著,文章闡述了通過實(shí)現(xiàn)聽覺模型反演過程從信號的自相關(guān)圖譜中恢復(fù)出原始的語音信號的過程。
The laboratory has proposed several speaker recognition methods involving computational auditory models , modular neural networks , gaussian mixed models , hidden markov models , and implemented a recognition framework combining semantic and voiceprint information 實(shí)驗(yàn)室提出了基于聽覺計(jì)算模型、模塊化神經(jīng)網(wǎng)絡(luò)、高斯混合模型、隱馬爾科夫模型等說話人識別方法,以及結(jié)合語義和聲紋信息的說話人識別框架。