Language Identification using Acoustic Models and Speaker Compensated Cepstral-Time Matrices | IEEE Conference Publication | IEEE Xplore