A voice recognition program that recognizes speakers' voice based on a model trained with speakers' voice.