Speech emotion recognition via multiple fusion under spatial–temporal parallel network | Publicación