2,753 Hits in 5.1 sec

On recognition of non-native speech using probabilistic lexical model

Marzieh Razavi, Mathew Magimai Doss
2014 Interspeech 2014   unpublished
In this paper, we investigate the role of different factors such as type of lexical model and choice of acoustic units in recognition of speech uttered by non-native speakers.  ...  Despite various advances in automatic speech recognition (ASR) technology, recognition of speech uttered by non-native speakers is still a challenging problem.  ...  Discussion and Conclusion In this paper, we studied the role of the probabilistic lexical model in the KL-HMM framework in improving non-native speech recognition without using any adaptation data.  ... 
doi:10.21437/interspeech.2014-6 fatcat:qtudkky7wjg3lnvkbt2xyh2o6a

Integrated pronunciation learning for automatic speech recognition using probabilistic lexical modeling

Ramya Rasipuram, Marzieh Razavi, Mathew Magimai-Doss
2015 2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)  
Standard automatic speech recognition (ASR) systems use phoneme-based pronunciation lexicon prepared by linguistic experts.  ...  In this paper, we propose a graphemebased ASR approach in the framework of probabilistic lexical modeling that integrates pronunciation learning as a stage in ASR system training, and exploits both acoustic  ...  We study multi-accent non-native speech recognition using the HIWIRE corpus [20] 2 .  ... 
doi:10.1109/icassp.2015.7178958 dblp:conf/icassp/RasipuramRM15 fatcat:jedikerrvzeilaay4ma6gl3rgm

On modeling context-dependent clustered states: Comparing HMM/GMM, hybrid HMM/ANN and KL-HMM approaches

Marzieh Razavi, Ramya Rasipuram, Mathew Magimai-Doss
2014 2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)  
Index Terms-HMM/GMM, hybrid HMM/ANN, Kullback-Leibler divergence based HMM, context-dependent subword units, non-native speech recognition  ...  The reduction in number of clustered states has broader implications on model complexity and data sparsity issues.  ...  WER on native speech decreases. On the other hand, as number of acoustic states is increased, the WER on non-native speech eventually increases for all the three systems.  ... 
doi:10.1109/icassp.2014.6855090 dblp:conf/icassp/RazaviRM14 fatcat:4gyzb2hozfhrtmgo4e3pg35xcq

Computer-assisted pronunciation training—Speech synthesis is almost all you need

Daniel Korzekwa, Jaime Lorenzo-Trueba, Thomas Drugman, Bozena Kostek
2022 Speech Communication  
Non-native English speech corpora of German, Italian, and Polish speakers are used in the evaluations.  ...  If we had a generative model that could mimic non-native speech and produce any amount of training data, then the task of detecting pronunciation errors would be much easier.  ...  [37] use a pair of speech signals from a student and a native speaker to classify native and non-native speech. Mauro et al.  ... 
doi:10.1016/j.specom.2022.06.003 fatcat:cl7ppxoegngc3bmuc5rzkw63mu

Probabilistic lexical modeling and unsupervised training for zero-resourced ASR

Ramya Rasipuram, Marzieh Razavi, Mathew Magimai-Doss
2013 2013 IEEE Workshop on Automatic Speech Recognition and Understanding  
Standard automatic speech recognition (ASR) systems rely on transcribed speech, language models, and pronunciation dictionaries to achieve state-of-the-art performance.  ...  In this paper, we propose a novel zero-resourced ASR approach to train acoustic models that only uses list of probable words from the language of interest.  ...  . • Non-native speech recognition [9] : Here, the goal was to build ASR system for non-native speech with limited acoustic resources.  ... 
doi:10.1109/asru.2013.6707771 dblp:conf/asru/RasipuramRM13 fatcat:2myb7plpxzbqpedt2hl6chzjlm

Regularized Speaker Adaptation of KL-HMM for Dysarthric Speech Recognition

Myungjong Kim, Younggwan Kim, Joohong Yoo, Jun Wang, Hoirin Kim
2017 IEEE transactions on neural systems and rehabilitation engineering  
Evaluation of the proposed speaker adaptation method on a database of several hundred words for 30 speakers consisting of 12 mildly dysarthric, 8 moderately dysarthric, and 10 non-dysarthric control speakers  ...  To capture the phonetic variation, Kullback-Leibler divergence based hidden Markov model (KL-HMM) is adopted, where the emission probability of state is parametrized by a categorical distribution using  ...  Acknowledgments This work was supported by the National Research Foundation of Korea (No. 2014r1a2a2a01007650) and the National Institutes of Health (No. r03dc013990).  ... 
doi:10.1109/tnsre.2017.2681691 pmid:28320669 pmcid:PMC5591083 fatcat:6c2betzlffhydnjfse6sdcl3qi

Speech Perception and Spoken Word Recognition: Past and Present

Peter W. Jusczyk, Paul A. Luce
2002 Ear and Hearing  
spoken word recognition, and research on how infants acquire the capacity to perceive their native language.  ...  Our foci in this review fall on three principle topics: early work on the discrimination and categorization of speech sounds, more recent efforts to understand the processes and representations that subserve  ...  ACKNOWLEDGMENTS: Preparation of the present manuscript was supported by a Senior Scientist Award from NIMH (01490) and a Research Grant from NICHD (15795) to PWJ and a Research Grant from NIDCD  ... 
doi:10.1097/00003446-200202000-00002 pmid:11881915 fatcat:2tj5axvwnrglzo2hmijp3z27vq

Multimodal Fusion of Multirate Acoustic, Prosodic, and Lexical Speaker Characteristics for Native Language Identification

Prashanth Gurunath Shivakumar, Sandeep Nallan Chakravarthula, Panayiotis Georgiou
2016 Interspeech 2016  
Native language identification from acoustic signals of L2 speakers can be useful in a range of applications such as informing automatic speech recognition (ASR), speaker recognition, and speech biometrics  ...  On the classification side we employ SVM, i-Vector, DNN and bottleneck features, and maximum-likelihood models.  ...  Identification of the native language (L1) of a non-native English speaker from English (L2) speech is a challenging research problem.  ... 
doi:10.21437/interspeech.2016-1312 dblp:conf/interspeech/ShivakumarCG16 fatcat:fg7orfwdfbgmdnaqdlbtrlnzia

Acoustic and lexical resource constrained ASR using language-independent acoustic model and language-dependent probabilistic lexical model

Ramya Rasipuram, Mathew Magimai-Doss
2015 Speech Communication  
The potential and the efficacy of the proposed approach is demonstrated through experiments and comparisons with other approaches on three different ASR tasks: non-native and accented speech recognition  ...  One of the key challenges involved in building statistical automatic speech recognition (ASR) systems is modeling the relationship between subword units or "lexical units" and acoustic feature observations  ...  (a) on the HIWIRE non-native accented speech recognition task, (b) on the Greek ASR task.  ... 
doi:10.1016/j.specom.2014.12.006 fatcat:hqyp5puyh5brjmatjv44z52ici

Pre-lexical abstraction of speech in the auditory cortex

Jonas Obleser, Frank Eisner
2009 Trends in Cognitive Sciences  
According to most cognitive models of spoken word recognition, this complexity is dealt with before lexical access via a process of abstraction from the acoustic signal to pre-lexical categories.  ...  Recent advances in animal neurophysiology and human functional imaging have made it possible to investigate the processing of speech in terms of probabilistic cortical maps rather than simple cognitive  ...  Acknowledgements J.O. is employed by the Max Planck Society, Germany, and received additional funding from the Landesstiftung Baden-Wü rttemberg gGmbH during initial phases of this project.  ... 
doi:10.1016/j.tics.2008.09.005 pmid:19070534 fatcat:mjfc24ww3benfgz4jm47ii6d2y

Page 486 of Behavior Research Methods Vol. 36, Issue 3 [page]

2004 Behavior Research Methods  
Frequency analysis of English usage: Lexicon and grammar. Boston: Houghton Mifflin. Gaygen, D. E. (1997). The effects of probabilistic phonotactics on the .segmentation of continuous .speech.  ...  The effect of probabilistic phonotactics on lexical acquisition. Clinical Linguistics & Phonetics, 14, 407-425. Treiman, R. (1986). The division between onsets and rimes in English syllables.  ... 

Managing Speech Perception Data Sets [chapter]

2022 The Open Handbook of Linguistic Data Management  
Acknowledgments The order of authorship is alphabetical. The research described in this contribution was financially supported by the Max Planck Society and all authors were previ-  ...  The principal stimulus for the project was to provideinput data for a probabilistic model of spoken-word recognition, the SHORTLIST-B model of Dutch spokenword recognition (Norris & McQueen 2008), and  ...  Although most citations are in publications on the central issue of native versus non-native speech perception, the study has also been cited on non-nativeness effects at higher levels of linguistic processing  ... 
doi:10.7551/mitpress/12200.003.0055 fatcat:a3l77jef75afxffurxlwowe5uy

Articulatory feature based continuous speech recognition using probabilistic lexical modeling

Ramya Rasipuram, Mathew Magimai.-Doss
2016 Computer Speech and Language  
., Articulatory feature based continuous speech recognition using probabilistic lexical modeling. Comput. Speech Lang. (2015), http://dx.  ...  -Doss, M., Articulatory feature based continuous speech recognition using probabilistic lexical modeling. Comput. Speech Lang. (2015) , http://dx.  ...  Acknowledgments This work was partly supported by the Swiss NSF through the grants "Flexible Grapheme-Based Automatic Speech Recognition (FlexASR, grant numbers 124985 and 146229)" and partly by the Commission  ... 
doi:10.1016/j.csl.2015.04.003 fatcat:6xv4ei5uuberdomxok4l6wsivy

Automatic Detection of Accent and Lexical Pronunciation Errors in Spontaneous Non-Native English Speech

Konstantinos Kyriakopoulos, Kate M. Knill, Mark J.F. Gales
2020 Interspeech 2020  
Three annotated corpora of non-native English speech by speakers of multiple L1s are analysed, the consistency of human annotation investigated and a method presented for detecting individual accent and  ...  Detecting individual pronunciation errors and diagnosing pronunciation error tendencies in a language learner based on their speech are important components of computer-aided language learning (CALL).  ...  Instead, an ASR system trained on non-native learners of English is used [36, 37] .  ... 
doi:10.21437/interspeech.2020-2881 dblp:conf/interspeech/KyriakopoulosKG20 fatcat:vfsdev64cza5bou2b2uv5mjhr4

Spoken word recognition in quiet and noise by native and non‐native listeners: Effects of age of immersion and vocabulary size

Astrid Z. Doty, Catherine L. Rogers, Judith B. Bryant
2009 Journal of the Acoustical Society of America  
to Mom and Dad, your desire for my success and happiness made me strive for both; to Kurt, Heidi, Christine, Michele, and Eric for your encouragement and love. to you all, I give my deepest expression of  ...  Would recognition of the word be facilitated from its probabilistic phonotactics in ways that would not benefit a native listener because of the native listener's lexical focus?  ...  The facilitative effects of probabilistic phonotactics for non-words occur because non-words fail to activate competing lexical representations.  ... 
doi:10.1121/1.4784690 fatcat:qhvgazsc7zb2nomhelp6tlhpru
« Previous Showing results 1 — 15 out of 2,753 results