Applying Emotional Factor Analysis and I-Vector to Emotional Speaker Recognition

被引：0

作者：

Chen, Li ^{[1
]}

Yang, Yingchun ^{[1
]}

机构：

[1] Zhejiang Univ, Coll Comp Sci & Technol, Hangzhou 310003, Zhejiang, Peoples R China

来源：

BIOMETRIC RECOGNITION: CCBR 2011 | 2011年 / 7098卷

关键词：

EFA; I-Vector; emotional speaker recognition;

D O I：

暂无

中图分类号：

TP301 [理论、方法];

学科分类号：

081202 ;

摘要：

Emotion variability is an important factor that degrades the performce of speaker recognition system. This paper borrows ideas from Joint Factor Analysis (JFA) algorithm based on the similarity between emotion effect and channel effect and develops Emotional Factor Analysis (EFA) into solving the emotion variability problem. 1-Vector is appiled also. The experiment carried on MASC (Madarin Affective Speech Corpus) shows that EFA and I-Vector method can bring an IR increase of 7%similar to 10% and an EER reduction of 3%similar to 4% compared with the GMM-UBM system.

引用

页码：174 / 179

页数：6

共 7 条

[1]

[Anonymous], 2007, INTERSPEECH

[2]

[Anonymous], 2006, ODYSSEY 2006

[3] Front-End Factor Analysis for Speaker Verification [J].

Dehak, Najim ;

Kenny, Patrick J. ;

Dehak, Reda ;

Dumouchel, Pierre ;

Ouellet, Pierre .

IEEE TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING, 2011, 19 (04) :788-798

[4]

Dehak N, 2009, INTERSPEECH 2009: 10TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION 2009, VOLS 1-5, P1527

[5] Joint factor analysis versus eigenchannels in speaker recognition [J].

Kenny, Patrick ;

Boulianne, Gilles ;

Ouellet, Pierre ;

Dumouchel, Pierre .

IEEE TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING, 2007, 15 (04) :1435-1447

[6] Speaker and session variability in GMM-based speaker verification [J].

Kenny, Patrick ;

Boulianne, Gilles ;

Ouellet, Pierre ;

Dumouchel, Pierre .

IEEE TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING, 2007, 15 (04) :1448-1460

[7]

Shriberg E., 2009, INT 2009 BRIGHT UK, P1551

← 1 →