ISCA Archive Interspeech 2009
ISCA Archive Interspeech 2009

Acoustic modeling using exponential families

Vaibhava Goel, Peder A. Olsen

We present a framework to utilize general exponential families for acoustic modeling. Maximum Likelihood (ML) parameter estimation is carried out using sampling based estimates of the partition function and expected feature vector. Markov Chain Monte Carlo procedures are used to draw samples from general exponential densities. We apply our ML estimation framework to two new exponential families to demonstrate the modeling flexibility afforded by this framework.


doi: 10.21437/Interspeech.2009-437

Cite as: Goel, V., Olsen, P.A. (2009) Acoustic modeling using exponential families. Proc. Interspeech 2009, 1423-1426, doi: 10.21437/Interspeech.2009-437

@inproceedings{goel09_interspeech,
  author={Vaibhava Goel and Peder A. Olsen},
  title={{Acoustic modeling using exponential families}},
  year=2009,
  booktitle={Proc. Interspeech 2009},
  pages={1423--1426},
  doi={10.21437/Interspeech.2009-437}
}