A Study of Imposture on HMM-based Speaker Verification Systems Using Synthetic Speech

17 References
  • HITOTSUMATSU Takafumi
    Interdisciplinary Graduate School of Science and Engineering, Tokyo Institute of Technology
  • MASUKO Takashi
    Interdisciplinary Graduate School of Science and Engineering, Tokyo Institute of Technology
  • KOBAYASHI Takao
    Interdisciplinary Graduate School of Science and Engineering, Tokyo Institute of Technology
  • TOKUDA Keiichi
    Department of Computer Science, Nagoya Institute of Technology

Bibliographic Information

Other Title
  • HMMに基づく話者照合システムにおける合成音声による詐称の検討

Search this article

Description

Speaker verification is the technique to judge whether input speech is the same at the claimed speaker's speech. The text-prompted speaker verification techniques are robust to the impostor with playing back recorded voice of a registered speaker. In this paper, we show that advances on text-to-speech synthesis techniques can defeat HMM-based text-prompted speaker verification systems. We construct a HMM-based text-prompted speaker verification system. Then we investigate whether synthetic speech generated from HMM-based speech synthesis system is accepted by the system. Moreover synthetic speech generated from speaker independent models adapted by a speaker adaptation technique is tested.

Journal

  • IPSJ SIG Notes

    IPSJ SIG Notes 24 75-82, 1998-12-10

    Information Processing Society of Japan (IPSJ)

References(17)*help

See more

Keywords

Details 詳細情報について

  • CRID
    1574231876805582720
  • NII Article ID
    110002948720
  • NII Book ID
    AN10442647
  • ISSN
    09196072
  • Text Lang
    ja
  • Data Source
    • CiNii Articles

Report a problem

Back to top