Document Type

Conference Proceeding

Language

eng

Format of Original

4 p.

Publication Date

7-2013

Publisher

Institute of Electrical and Electronics Engineers

Source Publication

2013 IEEE China Summit & International Conference on Signal and Information Processing (ChinaSIP)

Original Item ID

10.1109/ChinaSIP.2013.6625321

Abstract

This paper introduces the use of two new features for speaker identification, Residual Phase Cepstrum Coefficients (RPCC) and Glottal Flow Cepstrum Coefficients (GLFCC), to capture speaker-specific characteristics from their vocal excitation patterns. Results on a cross-lingual speaker identification task taken from the NIST 2004 SRE demonstrate that these RPCC and GLFCC features are significantly more accurate than traditional melfrequency cepstral coefficients (MFCC). In particular, these two new features give better results with smaller amounts of training data, due to lower model complexity.

Comments

Accepted version. Published as part of the proceedings of the conference, 2013 IEEE China Summit & International Conference on Signal and Information Processing (ChinaSIP), 2013: 170-173. DOI. © 2013 Institute of Electrical and Electronics Engineers. Used with permission.

johnson_5043acc.docx (223 kB)
ADA Accessible Version

Share

COinS