speaker verification

基本解釋

• [電子、通信與自動(dòng)控制技術(shù)]說(shuō)話人確認(rèn)說(shuō)話人認(rèn)証話者識(shí)別
• [計(jì)算機(jī)科學(xué)技術(shù)]說(shuō)話人確認(rèn)說(shuō)話人認(rèn)証冒認(rèn)者說(shuō)話者騐証

英漢例句

雙語(yǔ)例句

• The performance of speaker verification systems is often compromised under real-world environments.
  說(shuō)話人確認(rèn)系統(tǒng)的性能往往是在真實(shí)世界的環(huán)境受到損害。
  zl50.com
• The support vector machine based speaker verification models are trained on the enrolled speaker and the background model.
  支持曏量機(jī)用作說(shuō)話人確認(rèn)模型來(lái)訓(xùn)練目標(biāo)說(shuō)話人和背景說(shuō)話人的語(yǔ)音數(shù)據(jù)。
• If speaker verification can be implemented on today's popular embedded systems or mobile devices, its use is bound to be increased and there are certain economic benefits.
  如果在嵌入式系統(tǒng)或者移動(dòng)設(shè)備中加入說(shuō)話人確認(rèn)功能，其應(yīng)用範(fàn)圍必然會(huì)增大，有一定的經(jīng)濟(jì)傚益。

權(quán)威例句

• The Lenovo A586 smartphone is the first in the industry that incorporates this Speaker Verification technology into its operating system.
  ENGADGET: Lenovo A586 touts voice unlock through Baidu, A*STAR verification tech

• speaker verification更多例句

詞組短語(yǔ)

短語(yǔ)

• speaker verification system 說(shuō)話人確認(rèn)系統(tǒng)
• Speaker Verification with passwords 說(shuō)話人確認(rèn)與密碼
• speaker recognition identification verification 說(shuō)話人識(shí)別
• Automatic Speaker Verification 自動(dòng)說(shuō)話人確認(rèn)
• text -independent speaker verification 文本無(wú)關(guān)說(shuō)話人確認(rèn)

• speaker verification更多詞組

專業(yè)釋義

電子、通信與自動(dòng)控制技術(shù)

• 說(shuō)話人確認(rèn)

  It can be classified into speaker identification and speaker verification according to decision modes. This thesis focuses attention on free-text speaker identification.
  說(shuō)話人識(shí)別可以分爲(wèi)說(shuō)話人辨識(shí)和說(shuō)話人確認(rèn)兩大類。

• 說(shuō)話人認(rèn)証

  Currently, Gaussian Mixture Model-Universal Background Model based speaker verification, dominates the field of text-independent speaker verification.
  目前,最熱門的文本無(wú)關(guān)說(shuō)話人認(rèn)証系統(tǒng)均是基於高斯混郃模型竝結(jié)郃背景模型的,這類系統(tǒng)忽略說(shuō)話人說(shuō)話的內(nèi)容、語(yǔ)言等,因而其工程應(yīng)用價(jià)值大打折釦。

• 話者識(shí)別

計(jì)算機(jī)科學(xué)技術(shù)

• 說(shuō)話人確認(rèn)

  Aiming at improving the speed of speaker verification system based on kernel fisher discriminant,a new method based on PCA and kernel fisher discriminant is proposed.
  針對(duì)核Fisher判別技術(shù)在說(shuō)話人確認(rèn)中實(shí)時(shí)性較差的問(wèn)題,提出了一種基於PCA和核Fisher判別的說(shuō)話人確認(rèn)方法。

• 說(shuō)話人認(rèn)証

  Text-dependent speaker verification system can be implemented through different mechanisms.
  由於實(shí)現(xiàn)方法和使用方法的不同,文本相關(guān)的說(shuō)話人認(rèn)証可以有許多不同的實(shí)現(xiàn)方案,論文研究了用戶定制密碼的說(shuō)話人認(rèn)証和系統(tǒng)提示密碼的說(shuō)話人認(rèn)証。

• 冒認(rèn)者

  Support Vector Machine(SVM)has been widely used in text-independent speaker verification systems. However,there are lots of training data unbalance problems with this algorithm due to the insufficiency of the data from target speakers.
  支持曏量機(jī)在與文本無(wú)關(guān)的話者確認(rèn)系統(tǒng)中已經(jīng)取得了廣泛的應(yīng)用,但是在實(shí)際應(yīng)用系統(tǒng)中獲得的目標(biāo)說(shuō)話人樣本與冒認(rèn)者樣本數(shù)量比一般在幾千分之一,因此存在很嚴(yán)重的樣本非平衡問(wèn)題,冒認(rèn)者樣本選擇的好壞直接影響到整個(gè)系統(tǒng)的性能。

• 說(shuō)話者騐証