Text this: Secure speaker identification in open and closed environments modeled with symmetric comb filters.