Result Details

TOWARDS MACHINES THAT KNOW WHEN THEY DO NOT KNOW: SUMMARY OF WORK DONE AT 2014 FREDERICK JELINEK MEMORIAL WORKSHOP

HEŘMANSKÝ, H.; BURGET, L.; COHEN, J.; DUPOUX, E.; FELDMAN, N.; GODFREY, J.; KHUDANPUR, S.; MACIEJEWSKI, M.; MALLIDI, S.; MENON, A.; OGAWA, T.; PEDDINTI, V.; ROSE, R.; STERN, R.; WIESNER, M.; VESELÝ, K. TOWARDS MACHINES THAT KNOW WHEN THEY DO NOT KNOW: SUMMARY OF WORK DONE AT 2014 FREDERICK JELINEK MEMORIAL WORKSHOP. In Proceedings of 2015 IEEE International Conference on Acoustics, Speech and Signal Processing. South Brisbane, Queensland: IEEE Signal Processing Society, 2015. p. 5009-5013. ISBN: 978-1-4673-6997-8.
Type
conference paper
Language
English
Authors
Heřmanský Hynek, prof. Ing., Dr. Eng., DCGM (FIT)
Burget Lukáš, doc. Ing., Ph.D., DCGM (FIT)
Cohen Jordan
Dupoux Emmanuel
Feldman Naomi
Godfrey John
Khudanpur Sanjeev
Maciejewski Matthew
Mallidi Sri Harish, FIT (FIT)
Menon Anjali
Ogawa Tetsuji, DCGM (FIT)
Peddinti Vijayaditya
Rose Richard
Stern Richard
Wiesner Matthew, PhD.
Veselý Karel, Ing., Ph.D., DCGM (FIT)
Abstract

In this paper we have proposed and investigated several new techniques for predicting accuracy of estimation of posterior probabilities of speech sounds on previously unseen data, and have shown the feasibility of this task.

Keywords

Performance monitoring, confidenceestimation, multistream recognition of speech

URL
Annotation

A group of junior and senior researchers gathered as a part of the 2014 Frederick Jelinek Memorial Workshop in Prague to address the problem of predicting the accuracy of a nonlinear Deep Neural Network probability estimator for unknown data in a different application domain from the domain in which the estimator was trained. The paper describes the problem and summarizes approaches that were taken by the group1

Published
2015
Pages
5009–5013
Proceedings
Proceedings of 2015 IEEE International Conference on Acoustics, Speech and Signal Processing
Conference
2015 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING (ICASSP)
ISBN
978-1-4673-6997-8
Publisher
IEEE Signal Processing Society
Place
South Brisbane, Queensland
DOI
10.1109/ICASSP.2015.7178924
UT WoS
000427402905025
EID Scopus
2-s2.0-84946071181
BibTeX 
@inproceedings{BUT119897,
  author="Hynek {Heřmanský} and Lukáš {Burget} and Jordan {Cohen} and Emmanuel {Dupoux} and Naomi {Feldman} and John {Godfrey} and Sanjeev {Khudanpur} and Matthew {Maciejewski} and Sri Harish {Mallidi} and Anjali {Menon} and Tetsuji {Ogawa} and Vijayaditya {Peddinti} and Richard {Rose} and Richard {Stern} and Matthew {Wiesner} and Karel {Veselý}",
  title="TOWARDS MACHINES THAT KNOW WHEN THEY DO NOT KNOW: SUMMARY OF WORK DONE AT 2014 FREDERICK JELINEK MEMORIAL WORKSHOP",
  booktitle="Proceedings of 2015 IEEE International Conference on Acoustics, Speech and Signal Processing",
  year="2015",
  pages="5009--5013",
  publisher="IEEE Signal Processing Society",
  address="South Brisbane, Queensland",
  doi="10.1109/ICASSP.2015.7178924",
  isbn="978-1-4673-6997-8",
  url="https://ieeexplore.ieee.org/document/7178924"
}
Files
Projects
IARPA Building Speech Recognition for Keyword Search in a New Language in a Week with Limited Training Data (BABEL) - Babelon, BBN, start: 2012-03-05, end: 2016-11-04, completed
Meeting Assistant (MINT), TAČR, Program aplikovaného výzkumu a experimentálního vývoje ALFA, TA04011311, start: 2014-10-01, end: 2017-12-31, completed
Research groups
Speech Data Mining Research Group BUT Speech@FIT (RG SPEECH)
Departments
Department of Computer Graphics and Multimedia (DCGM)
Back to top