Language Identification using Acoustic Models and Speaker Compensated Cepstral-Time Matrices.
Fabio CastaldoEmanuele DalmassoPietro LafaceDaniele ColibroClaudio VairPublished in: ICASSP (4) (2007)
Keyphrases
- language identification
- acoustic models
- speech recognition
- speaker identification
- speaker independent
- broadcast news
- speech recognizer
- speech signal
- automatic speech recognition
- noisy environments
- hidden markov models
- speaker verification
- pattern recognition
- language model
- document images
- neural network
- speech recognition systems
- feature extraction
- gaussian mixture model
- web information retrieval
- training data