Advanced search
Add to list

Comprehensive integration of linguistic features in a human inspired speech recognition architecture

Claudia Matos Veliz (UGent) , Kris Demuynck (UGent) and Veronique Hoste (UGent)
Author
Organization
Abstract
Despite the fact that speech recognition involves a fair amount of natural language processing, there has been very little collaboration between linguists and speech recognition engineers. Nowadays, both domains use similar techniques such as log linear models and graphical models. Moreover, the advent of new powerful algorithms allows the deployment of probabilistic models that are no longer limited by the oversimpli cations present in most of today’s speech recognition systems and should allow successful integration of rich linguistic knowledge in automatic speech recognition systems. In this project, a novel speech recognition framework and a set of accompanying, contemporary linguistic models is co-developed. The envisioned layered architecture combines local inference with inter-layer message passing, similar to what is believed to underlay human speech recognition. Compared to the current architectures, such layer-wise structure facilitates the incorporation of additional knowledge (linguistic or other), and hence is better equipped to model the nenuances and complex inter-dependencies in speech.
Keywords
LT3

Citation

Please use this url to cite or link to this publication:

MLA
Matos Veliz, Claudia, Kris Demuynck, and Veronique Hoste. “Comprehensive Integration of Linguistic Features in a Human Inspired Speech Recognition Architecture.” FEARS 2019 : Book of Abstracts. Ed. Kevin Dekemele. Ghent University, 2019. 90–91. Print.
APA
Matos Veliz, C., Demuynck, K., & Hoste, V. (2019). Comprehensive integration of linguistic features in a human inspired speech recognition architecture. In K. Dekemele (Ed.), FEARS 2019 : book of abstracts (pp. 90–91). Presented at the 19th FEA Research Symposium, FEARS 2019, Ghent University.
Chicago author-date
Matos Veliz, Claudia, Kris Demuynck, and Veronique Hoste. 2019. “Comprehensive Integration of Linguistic Features in a Human Inspired Speech Recognition Architecture.” In FEARS 2019 : Book of Abstracts, ed. Kevin Dekemele, 90–91. Ghent University.
Chicago author-date (all authors)
Matos Veliz, Claudia, Kris Demuynck, and Veronique Hoste. 2019. “Comprehensive Integration of Linguistic Features in a Human Inspired Speech Recognition Architecture.” In FEARS 2019 : Book of Abstracts, ed. Kevin Dekemele, 90–91. Ghent University.
Vancouver
1.
Matos Veliz C, Demuynck K, Hoste V. Comprehensive integration of linguistic features in a human inspired speech recognition architecture. In: Dekemele K, editor. FEARS 2019 : book of abstracts. Ghent University; 2019. p. 90–1.
IEEE
[1]
C. Matos Veliz, K. Demuynck, and V. Hoste, “Comprehensive integration of linguistic features in a human inspired speech recognition architecture,” in FEARS 2019 : book of abstracts, Ghent, 2019, pp. 90–91.
@inproceedings{8609701,
  abstract     = {Despite the fact that speech recognition involves a fair amount of natural language processing, there has been very little collaboration between linguists and speech recognition engineers. Nowadays, both domains use similar techniques such as log linear models and graphical models. Moreover, the advent of new powerful algorithms allows the deployment of probabilistic models that are no longer limited by the oversimpli cations present in most of today’s speech recognition systems and should allow successful integration of rich linguistic knowledge in automatic speech recognition systems. In this project, a novel speech recognition framework and a set of accompanying, contemporary linguistic models is co-developed. The envisioned layered architecture combines local inference with inter-layer message passing, similar to what is believed to underlay human speech recognition. Compared to the current architectures, such layer-wise structure facilitates the incorporation of additional knowledge (linguistic or other), and hence is better equipped to model the  nenuances and complex inter-dependencies in speech.},
  author       = {Matos Veliz, Claudia and Demuynck, Kris and Hoste, Veronique},
  booktitle    = {FEARS 2019 : book of abstracts},
  editor       = {Dekemele, Kevin},
  isbn         = {NA},
  keywords     = {LT3},
  language     = {eng},
  location     = {Ghent},
  pages        = {90--91},
  publisher    = {Ghent University},
  title        = {Comprehensive integration of linguistic features in a human inspired speech recognition architecture},
  url          = {http://www.fears.ugent.be/booklet.pdf},
  year         = {2019},
}