Hybrid segmental-LVQ/HMM for large vocabulary speech recognition | IEEE Conference Publication | IEEE Xplore