Cargando…
Multi-Path and Group-Loss-Based Network for Speech Emotion Recognition in Multi-Domain Datasets
Speech emotion recognition (SER) is a natural method of recognizing individual emotions in everyday life. To distribute SER models to real-world applications, some key challenges must be overcome, such as the lack of datasets tagged with emotion labels and the weak generalization of the SER model fo...
Autores principales: | , , , , , , |
---|---|
Formato: | Online Artículo Texto |
Lenguaje: | English |
Publicado: |
MDPI
2021
|
Materias: | |
Acceso en línea: | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7956608/ https://www.ncbi.nlm.nih.gov/pubmed/33668254 http://dx.doi.org/10.3390/s21051579 |