Compensate multiple distortions for speaker recognition systems

Mohammad Mohammadamini (1) , Driss Matrouf (1) , Jean-Francois Bonastre (1) , Romain Serizel (2) , Sandipana Dowerah (2) , Denis Jouvet (2)

1 LIA - Laboratoire Informatique d'Avignon
2 MULTISPEECH - Speech Modeling for Facilitating Oral-Based Communication

The performance of speaker recognition systems reduces dramatically in severe conditions in the presence of additive noise and/or reverberation. In some cases, there is only one kind of domain mismatch like additive noise or reverberation, but in many cases, there are more than one distortion. Finding a solution for domain adaptation in the presence of different distortions is a challenge. In this paper we investigate the situation in which there is none, one or more of the following distortions: early reverberation, full reverberation, additive noise. We propose two configurations to compensate for these distortions. In the first one a specific denoising autoencoder is used for each distortion. In the second configuration, a denoising autoencoder is used to compensate for all of these distortions simultaneously. Our experiments show that, in the coexistence of noise and reverberation, the second configuration gives better results. For example, with the second configuration we obtained 76.6% relative improvement of EER for utterances longer than 12 seconds. For other situations in the presence of only one distortion, the second configuration gives almost the same results achieved by using a specific model for each distortion.

Mots clés: additive noise; early reverberation; full reverberation; x-vector; denoising autoencoder

Domaines: Intelligence artificielle [cs.AI]

Mohammad Mohammadamini, Driss Matrouf, Jean-Francois Bonastre, Romain Serizel, Sandipana Dowerah, et al.. Compensate multiple distortions for speaker recognition systems. EUSIPCO 2021 - 29th European Signal Processing Conference, Aug 2021, Dublin / Virtual, Ireland. ⟨10.23919/EUSIPCO54536.2021.9615983⟩. ⟨hal-03224675⟩