Unsupervised noisy environment adaptation algorithm using MLLR and speaker selection.
Miichi YamadaAkira BabaShinichi YoshizawaYuichiro MeraAkinobu LeeHiroshi SaruwatariKiyohiro ShikanoPublished in: INTERSPEECH (2001)
Keyphrases
- selection algorithm
- learning algorithm
- optimal solution
- dynamic programming
- detection algorithm
- computational complexity
- objective function
- np hard
- speech recognition
- segmentation algorithm
- fine tuning
- noisy environments
- probabilistic model
- unsupervised learning
- supervised learning
- semi supervised
- k means
- pattern recognition
- information retrieval