A dynamic approach to the selection of high order n-grams in phonotactic language recognition.
Mikel PeñagarikanoAmparo VaronaLuis Javier Rodríguez-FuentesGermán BordelPublished in: ICASSP (2011)
Keyphrases
- high order
- n gram
- higher order
- language model
- character n grams
- language specific
- pairwise
- low order
- bag of words
- text classification
- language independent
- feature extraction
- variable length
- markov random field
- language modeling
- part of speech
- tensor analysis
- action recognition
- lower order
- similarity measure
- computer vision
- bayesian logistic regression
- inside outside algorithm
- anomaly detection