Multilingual phone models for vocabulary-independent speech recognition tasks
Speech CommunicationPublished 1 August 2001
Joachim Köhler
Citations53
SJR quartileQ1
SJR score0.49
SNIP1.25
Generate an AI Snapshot to get a quick, structured summary of this paper.
Study Snapshot
ObjectiveStudy objective
MethodsResearch methodology
PopulationPopulation studied
Sample sizeSample sizes
OutcomesStudy outcomes here
ResultsStudy results comes here
LimitationsResearch study limitations comes here
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
Three different methods to develop multilingual phone models for flexible speech recognition tasks are presented and a huge reduction of the number of densities in the multilingual system is observed.
Abstract
S.21-30
Keywords
Computer Science
Language<b>The sounds of the world’s languages.</b> By Peter Ladefoged and Ian Maddieson. Oxford Cambridge, MA: Blackwell, 1996. Pp. xxi, 426. Paper $31.95.
1,740 Citations1998Geoffrey S. Nathan
AT&T Technical JournalA Probabilistic Distance Measure for Hidden Markov Models
384 Citations1985B.-H. Juang, L. R. Rabiner
A probabilistic distance measure for measuring the dissimilarity between pairs of hidden Markov models with arbitrary observation densities is proposed based on the Kullback-Leibler number and is consistent with the reestimation technique for hidden MarkOV models.
Speech CommunicationMultilingual spoken-language understanding in the MIT Voyager system
104 Citations1995James Glass, Giovanni Flammia +6 more
The description will focus on the development of the multilingual MIT Voyager spoken language system, which can engage in verbal dialogues with users about a geographical region within Cambridge, MA in the USA.
Multi-lingual phoneme recognition exploiting acoustic-phonetic similarities of sounds
93 Citations2002J. Köhler
A statistical distance measure is introduced to determine the similarities of sounds and a new method of modelling multi-lingual phonemes, which can be used for a variety of languages, is presented, to exploit the acoustic-phonetic similarities between several languages.
Fast bootstrapping of LVCSR systems with multilingual phoneme sets
86 Citations1997Tanja Schultz, Alex Waibel
An e cient method to bootstrap continuously spoken, large vocabulary speech recognition systems by multilingual phoneme sets by MULTI is described and achieves 100% language identi cation rate.
Calculation of distance measures between hidden Markov models
59 Citations1995M. Falkhausen, Herbert Reininger +1 more
Two methods to define a distance measure between any pair of Hidden Markov Models are investigated, the geometricaly motivated euclidean distance which solely incorporates the feature probabilities and the Kulback-Liebler distance which is based on the discriminating power of the probability measure on the space of feature sequences induced by the HMMs.
Language adaptation of multilingual phone models for vocabulary independent speech recognition tasks
55 Citations2002J. Köhler
Adaptation techniques for cross-language transfer showed that only 100 utterances from a new language were needed for adaptation and the recognition rate was improved from 79.9% to 84.3% using the MAP algorithm.
IEEE International Conference on Acoustics Speech and Signal ProcessingCross-lingual experiments with phone recognition
49 Citations1993Lori Lamel, J.-L. Gauvain
Research on speaker-independent continuous phone recognition for both French and English is presented and it is found that French is easier to recognize at the phone level (the phone error for BREF is 23.6% vs. 30.1% for WSJ), but harder to recognizing at the lexical level due to the larger number of homophones.
In-service adaptation of multilingual hidden-Markov-models
30 Citations2002Udo Bub, J. Köhler +1 more
A multilingual romanic/germanic seed model for a slavic target task and in tests on Slovene digits multilingual modeling yields the best recognition accuracy compared to other language dependent models.
Issues in large vocabulary, multilingual speech recognition
30 Citations1995Lori Lamel, Martine Adda‐Decker +1 more
The existing recognizer for American English and French, has been ported to British English and German and has been assessed in the context of the LRE SQALE project whose objective was to experiment with installing in Europe a multilingual evaluation paradigm for the assessment of large vocabulary, continuous speech recognition systems.
Concise Compendium of the World's Languages
24 Citations2003George Campbell
A model distance measure for talker clustering and identification
15 Citations2002Jonathan Foote, H.F. Silverman
This paper describes methods of talker clustering and identification based on a "distance" metric between discrete HMM output probabilities derived on a tree-based MMI partition of the feature space, rather than the usual vector quantization.
OHSU Digital CommonsAutomatic language identification with sequences of language-independent phoneme clusters
10 Citations1996Kay Berkling
A "word-spotting" algorithm is introduced motivated by a perceptual study reporting that human subjects use language-dependent phonemes and short sequences to identify languages and a mathematical model of the discrimination between two languages is developed.
Language-identification using language-dependent phonemes and language-independent speech units
7 Citations2002Paul Dalsgaard, Ove Andersen +2 more
Results from a number of experiments show that average language identification scores of close to 90% can be retained by the LID system presented here, even for a high number of language independent speech units.
Journal of the International Phonetic AssociationIPA volume 23 issue 2 Cover and Back matter
2 Citations1993
16-bit Sound I/O: Power Macintoshes and AVMacIntoshes, AudioMedia, any Apple-compatible driver 8-bitsound Reproduction: Apple integrated, MacRecorder Sound Reproduction, automatic trimming, manual scoring, text editing for scoring results, horizontal and vertical signal zoom