Přejít na obsah

Publikace

Detail publikace

Citace

Jan Zelinka and Jan Trmal and Luděk Müller : On Context-Dependent Neural Networks and Speaker Adaptation . IEEE 11th International Conference on Signal Processing, 2012.

Abstrakt

This paper describes evaluation of a neural network based hybrid LVCSR system. The~novelty of the evaluated hybrid system lies in speaker adaptation techniques that are employed to increase performance of neural networks for context-dependent phonetic units modeling. The performance comparison is done as follows. First, performances of different hybrid systems employing either a context-independent neural network or a context-dependent neural network are compared. Second, the influence of the recently published speaker adaptation technique called MELT is evaluated. Furthermore, several possible approaches to conversion of posterior probabilities into observation likelihoods, which are necessary for a hybrid LVSCR systems, are described and discussed in this paper.

Detail publikace

Název: On Context-Dependent Neural Networks and Speaker Adaptation
Autor: Jan Zelinka ; Jan Trmal ; Luděk Müller
Jazyk publikace: anglicky
Rok vydání: 2012
Typ publikace: Stať ve sborníku
Název časopisu / knihy: IEEE 11th International Conference on Signal Processing
ISBN: 978-1-4673-2194-5
/ 2019-11-18 12:57:01 /

BibTeX

@MISC{JanZelinka_2012_OnContext-Dependent,
 author = {Jan Zelinka and Jan Trmal and Lud\v{e}k M\"{u}ller},
 title = {On Context-Dependent Neural Networks and Speaker Adaptation},
 year = {2012},
 journal = {IEEE 11th International Conference on Signal Processing},
 ISBN = {978-1-4673-2194-5},
 url = {http://www.kky.zcu.cz/en/publications/JanZelinka_2012_OnContext-Dependent},
}