Using Audio Events to Extend a Multi-modal Public Speaking Database with Reinterpreted Emotional Annotations | Publicación