Using SVMS and discriminative models for speech recognition | IEEE Conference Publication | IEEE Xplore