A comparison of front-end configurations for robust speech recognition | IEEE Conference Publication | IEEE Xplore