Толық нәтижелер

AbstractThe current practice in virtual human dialogue systems is to use professional human recordings or limited-domain speech synthesis. Both approaches lead to good performance but at a high cost. To …...