Sciweavers

ICPR
2010
IEEE

Learning Virtual HD Model for Bi-model Emotional Speaker Recognition

13 years 1 months ago
Learning Virtual HD Model for Bi-model Emotional Speaker Recognition
Pitch mismatch between training and testing is one of the important factors causing the performance degradation of the speaker recognition system. In this paper, we adopted the missing feature theory and specified the Unreliable Region (UR) as the parts of the utterance with high emotioninduced pitch variation. To model these regions, a virtual HD (High Different from neutral, with large pitch offset) model for each target speaker was built from the virtual speech, which were converted from the neutral speech by the Pitch Transformation Algorithm (PTA). In the PTA, a polynomial transformation function was learned to model the relationship of the average pitch between the neutral and the high-pitched utterances. Compared with traditional GMM-UBM and our
Ting Huang, Yingchun Yang
Added 12 Feb 2011
Updated 12 Feb 2011
Type Journal
Year 2010
Where ICPR
Authors Ting Huang, Yingchun Yang
Comments (0)