Meta-aprendizado para a verificação de falante com áudios de curta duração
Wedi'i Gadw mewn:
| Prif Awdur: | |
|---|---|
| Dyddiad Cyhoeddi: | 2022 |
| Fformat: | Master thesis |
| Iaith: | por |
| Ffynhonnell: | Repositório Institucional da UFG |
| Download full: | http://repositorio.bc.ufg.br/tede/handle/tede/12072 |
Crynodeb: | In practical scenarios, a speaker verification model system must be able to identify a person given audios of any durations. However, existing speaker verification systems have low performance when dealing with short-length audios. To face this problem, the MLVL (Meta-Learning Variable-Length) approach was proposed, which consists of using audios with different durations within the same episode in the meta-learning of a prototypical network. The objective is to become text-independent speaker verification more robust to the context in which the verification audio is short-length. Models trained with the MLVL approach were evaluated in three different scenarios of short-length audios, obtaining 2.55% as the lowest EER (Equal Error Rate) value. Evaluating such models in audios with longer durations, the lowest EER value obtained was 2.40%. The results surpassed those obtained by several studies in the same scenarios, demonstrating the potential practical application of the proposed MLVL approach in a voice biometrics system. |
Eitemau Tebyg: Meta-aprendizado para a verificação de falante com áudios de curta duração
- Redes neurais profundas com fusão de características na verificação biométrica pela voz
- Verificação de texto manuscrito em dispositivos móveis
- Meta-research in aquatic ecology
- Motivação e esporte : uma intervenção das metas de realização em jovens atletas
- Professional fulfillment : concept and instrument proposition
- Como características metodológicas afetam a estimativa da razão sexual em peixes adultos?