Source and system features for speaker recognition using AANN models | IEEE Conference Publication | IEEE Xplore