Logo des Repositoriums

Auflistung nach Schlagwort "speaker verification"

1 - 2 von 2
  • Textdokument
    Impact of bandwidth and channel variation on presentation attack detection for speaker verification
    (BIOSIG 2017, 2017) Delgado,Héctor; Todisco,Massimiliano; Evans,Nicholas; Sahidullah,Md; Liu,Wei Ming; Alegre,Federico; Kinnunen,Tomi; Fauve,Benoit
    Vulnerabilities to presentation attacks can undermine confidence in automatic speaker verification (ASV) technology. While efforts to develop countermeasures, known as presentation attack detection (PAD) systems, are now under way, the majority of past work has been performed with high-quality speech data. Many practical ASV applications are narrowband and encompass various coding and other channel effects. PAD performance is largely untested in such scenarios. This paper reports an assessment of the impact of bandwidth and channel variation on PAD performance. Assessments using two current PAD solutions and two standard databases show that they provoke significant degradations in performance. Encouragingly, relative performance improvements of 98% can nonetheless be achieved through feature optimisation. This performance gain is achieved by optimising the spectro-temporal decomposition in the feature extraction process to compensate for narrowband speech. However, compensating for channel variation is considerably more challenging.
  • Konferenzbeitrag
    Improving channel robustness in text-independent speaker verification using adaptive virtual cohort models
    (BIOSIG 2012, 2012) Nautsch, Andreas; Schönwandt, Anne; Kasper, Klaus; Reininger, Herbert; Wagner, Martin
    In speaker verification, score normalization methods are a common practice to gain better performance and robustness. One kind of score normalization is cohort normalization, which uses information about the score behaviour of known impostors. During enrolment, impostor verifications are simulated to get a speaker-specific set of the most competitive impostors (the cohort). In the present paper, one virtual cohort speaker is synthesized using the most competitive impostor's Hidden Markov Models (HMMs). These impostors are also users of the system and therefore their models have channel-specific information contrary to the universal background model, which provides channeland speaker-independent models. On verification, cohort scores are obtained by an additional verification of the virtual cohort speaker. The cohort scores evaluate the candidate as an impostor. A cohort normalized score promises greater robustness. This paper will study the effect of the introduced cohort normalization technique on the speaker verification system atip VoxGuard, which is based on mel-frequency cepstral coefficients and HMMs. VoxGuard can be used as either a text-dependent or a text-independent verification system. In this paper, emphasis is placed on textindependent speaker verification. Experiments using the atip speech corpus and the SieTill speech corpus showed improvements measured by the equal error rate on performance and robustness.