Feature enhancement by speaker-normalized SPLICE for robust speech recognition

Yusuke Shinohara, Takashi Masuko, Masami Akamine

研究成果: Conference contribution

2 被引用数 (Scopus)

抄録

The SPLICE method of feature enhancement is known for its powerful performance. It learns a mapping from noisy to clean feature vectors given a set of stereo training data. However, feature vector variation caused by speaker changes conceals noise-induced variation, which is what we want to find in the SPLICE training. In this paper, an improvement of SPLICE by means of speaker-normalization is proposed. The training data is first normalized with respect to speaker variation, and a mapping is learned afterward. CMLLR with a GMM as its target is utilized for the speaker-normalization, where the GMM representing a standard speaker is learned via a novel variant of the speaker adaptive training. The proposed method was evaluated on Aurora2, and achieved a relative word error rate reduction of 38% over the conventional SPLICE.

本文言語English
ホスト出版物のタイトル2008 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP
ページ4881-4884
ページ数4
DOI
出版ステータスPublished - 2008 9月 16
外部発表はい
イベント2008 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP - Las Vegas, NV, United States
継続期間: 2008 3月 312008 4月 4

出版物シリーズ

名前ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
ISSN(印刷版)1520-6149

Other

Other2008 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP
国/地域United States
CityLas Vegas, NV
Period08/3/3108/4/4

ASJC Scopus subject areas

  • ソフトウェア
  • 信号処理
  • 電子工学および電気工学

フィンガープリント

「Feature enhancement by speaker-normalized SPLICE for robust speech recognition」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル