End-to-end speech emotion recognition using a novel context-stacking dilated convolution neural network

Amongst the various characteristics of a speech signal, the expression of emotion is one of the characteristics that exhibits the slowest temporal dynamics. Hence, a performant speech emotion recognition (SER) system requires a predictive model that is capable of learning sufficiently long temporal...

Descripción completa

Detalles Bibliográficos
Autores principales: Tang, Duowei, Kuppens, Peter, Geurts, Luc, van Waterschoot, Toon
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Springer International Publishing 2021
Materias:
Research
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8550764/
https://www.ncbi.nlm.nih.gov/pubmed/34721556
http://dx.doi.org/10.1186/s13636-021-00208-5

Internet

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8550764/
https://www.ncbi.nlm.nih.gov/pubmed/34721556
http://dx.doi.org/10.1186/s13636-021-00208-5

Ejemplares similares