Open AccessDiscriminative feature based on FWMW for playback speech detection
Author(s) -
Yang Jichen,
Liu Leian,
He Qianhua
Publication year - 2019
Publication title -
electronics letters
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.375
H-Index - 146
ISSN - 1350-911X
DOI - 10.1049/el.2019.1025
Subject(s) - discriminative model , speech recognition , computer science , pattern recognition (psychology) , feature (linguistics) , frame (networking) , artificial intelligence , voice activity detection , feature extraction , gaussian , word error rate , bin , segmentation , set (abstract data type) , speech processing , algorithm , programming language , telecommunications , philosophy , linguistics , physics , quantum mechanics
A discriminative feature extraction method for playback speech detection is proposed, it relies on the finding that frame‐wise magnitude‐spectrum weight (FWMW) can enlarge the difference between genuine speech and playback speech. The proposed FWMW is obtained by supplying the sum of every frame magnitude spectrum frequency bin as the weight on the frame magnitude spectrum. Then, a new feature based on FWMW is proposed, namely constant‐Q weight segmentation coefficients (CQWSCs). The experimental result on ASVspoof 2017 version 2.0 evaluation set using CQWSC indicates that: (i) FWMW can make the proposed feature has more discriminative ability and the equal error rate can decline 15.53 and 19.18% under Gaussian mixture model and deep neural network, respectively, and (ii) CQWSC performs better than some commonly used features.