The time sequence modeling ability of HMM and classifying ability of WNN were applied in the speaker recognition system independent from the text.

 
  • 該方法利用HMM的時(shí)序建模能力以及小波神經(jīng)網(wǎng)絡(luò )較強的模式分類(lèi)能力;進(jìn)行與文本無(wú)關(guān)的說(shuō)話(huà)人的識別.
今日熱詞
目錄 附錄 查詞歷史
国内精品美女A∨在线播放xuan