Human and Automatic Speaker Recognition over Telecommunication Channels

This work addresses the evaluation of the human and the automatic speaker recognition performances under different channel distortions caused by bandwidth limitation, codecs, and electro-acoustic user interfaces, among other impairments. Its main contribution is the demonstration of the benefits of communication channels of extended bandwidth, together with an insight into how speaker-specific characteristics of speech are preserved through different transmissions. It provides sufficient motivation for considering speaker recognition as a criterion for the migration from narrowband to enhanced bandwidths, such as wideband and super-wideband.

Addresses the evaluation of the human and the automatic performances under different channel distortions Demonstrates the benefits of communication channels of extended bandwidth Delivers an insight into how speaker-specific characteristics of speech are preserved through different transmissions Motivates speaker recognition as a criterion for the migration from narrowband to enhanced bandwidths, such as wideband and super-wideband Includes supplementary material: sn.pub/extras

Inhalt

Introduction.- Literature Review.- Human Speaker Identification Performance under Channel Degradations.- Importance of Intelligible Phonemes for Human Speaker Recognition in Different Bandwidths.- Automatic Speaker Verification Performance Under Channel Distortions.- Detecting Speaker-Discriminative Spectral Content in Wideband for Automatic Speaker Recognition.- Relations Among Speech Quality, Human Speaker Identification, and Automatic Speaker Verification. Conclusions and Future Work.