About: Confidence estimation, OOV detection and language ID using phone-to-word transduction and phone-level alignments

Facets (new session)
Description
Metadata
Settings
- owl:sameAs
- Inference Rule:

About: Confidence estimation, OOV detection and language ID using phone-to-word transduction and phone-level alignments Goto Sponge NotDistinct Permalink

An Entity of Type : http://linked.opendata.cz/ontology/domain/vavai/Vysledek, within Data Space : linked.opendata.cz associated with source document(s)

Attributes	Values
rdf:type	skos:Concept http://linked.opendata.cz/ontology/domain/vavai/Vysledek
Description	Automatic Speech Recognition (ASR) systems continue to make errors during search when handling various phenomena including noise, pronunciation variation, and out of vocabulary (OOV) words. Predicting the probability that a word is incorrect can prevent the error from propagating and perhaps allow the system to recover. This paper addresses the problem of detecting errors and OOVs for read Wall Street Journal speech when the word error rate (WER) is very low. It augments a traditional confidence estimate by introducing two novel methods: phone-level comparison using Multi-String Alignment (MSA) and word-level comparison using phone-to-word transduction. We show that features from phone and word string comparisons can be added to a standard maximum entropy framework thereby substantially improving performance in detecting both errors and OOVs. Additionally we show an extension to detecting English and accented English for the Language Identification (LID) task. Automatic Speech Recognition (ASR) systems continue to make errors during search when handling various phenomena including noise, pronunciation variation, and out of vocabulary (OOV) words. Predicting the probability that a word is incorrect can prevent the error from propagating and perhaps allow the system to recover. This paper addresses the problem of detecting errors and OOVs for read Wall Street Journal speech when the word error rate (WER) is very low. It augments a traditional confidence estimate by introducing two novel methods: phone-level comparison using Multi-String Alignment (MSA) and word-level comparison using phone-to-word transduction. We show that features from phone and word string comparisons can be added to a standard maximum entropy framework thereby substantially improving performance in detecting both errors and OOVs. Additionally we show an extension to detecting English and accented English for the Language Identification (LID) task. (en)
Title	Confidence estimation, OOV detection and language ID using phone-to-word transduction and phone-level alignments Confidence estimation, OOV detection and language ID using phone-to-word transduction and phone-level alignments (en)
skos:prefLabel	Confidence estimation, OOV detection and language ID using phone-to-word transduction and phone-level alignments Confidence estimation, OOV detection and language ID using phone-to-word transduction and phone-level alignments (en)
skos:notation	RIV/00216305:26230/08:PU76780!RIV10-MSM-26230___
http://linked.open...avai/riv/aktivita	Z
http://linked.open...avai/riv/aktivity	Z(MSM0021630528)
http://linked.open...vai/riv/dodaniDat	2010
http://linked.open...aciTvurceVysledku	Burget, Lukáš Schwarz, Petr Heřmanský, Hynek
http://linked.open.../riv/druhVysledku	D - Článek ve sborníku
http://linked.open...iv/duvernostUdaju	S - Úplné a pravdivé údaje nepodléhající ochraně podle zvláštních právních předpisů
http://linked.open...titaPredkladatele	Vysoké učení technické v Brně / Fakulta informačních technologií
http://linked.open...dnocenehoVysledku	361006
http://linked.open...ai/riv/idVysledku	RIV/00216305:26230/08:PU76780
http://linked.open...riv/jazykVysledku	eng - angličtina
http://linked.open.../riv/klicovaSlova	speech recognition<br> (en)
http://linked.open.../riv/klicoveSlovo	speech recognition<br>
http://linked.open...ontrolniKodProRIV	[475C15036C86]
http://linked.open...v/mistoKonaniAkce	Las Vegas
http://linked.open...i/riv/mistoVydani	Las Vegas
http://linked.open...i/riv/nazevZdroje	Proc. 2008 IEEE International Conference on Acoustics, Speech, and Signal Processing
http://linked.open...in/vavai/riv/obor	JC
http://linked.open...ichTvurcuVysledku	3 (xsd:int)
http://linked.open...cetTvurcuVysledku	5 (xsd:int)
http://linked.open...UplatneniVysledku	2008
http://linked.open...iv/tvurceVysledku	Burget, Lukáš Schwarz, Petr Zweig, Geoffrey Heřmanský, Hynek White, Christopher
http://linked.open...vavai/riv/typAkce	WRD - Světová
http://linked.open.../riv/zahajeniAkce	2008-03-30 (xsd:date)
http://linked.open...n/vavai/riv/zamer	Výzkum informačních technologií z hlediska bezpečnosti
number of pages	4 (xsd:int)
http://purl.org/ne...btex#hasPublisher	IEEE Signal Processing Society
https://schema.org/isbn	1-4244-1484-9
http://localhost/t...ganizacniJednotka	26230

Faceted Search & Find service v1.16.118 as of Jun 21 2024

Alternative Linked Data Documents: ODE Content Formats:

RDF

ODATA

Microdata

About

OpenLink Virtuoso version 07.20.3240 as of Jun 21 2024, on Linux (x86_64-pc-linux-gnu), Single-Server Edition (126 GB total memory, 112 GB memory in use)
Data on this page belongs to its respective rights holders.
Virtuoso Faceted Browser Copyright © 2009-2024 OpenLink Software