Network


Latest external collaboration on country level. Dive into details by clicking on the dots.

Hotspot


Dive into the research topics where Yannick Marchand is active.

Publication


Featured researches published by Yannick Marchand.


Computational Linguistics | 2000

A multistrategy approach to improving pronunciation by analogy

Yannick Marchand; Robert I. Damper

Pronunciation by analogy (PbA) is a data-driven method for relating letters to sound, with potential application to next-generation text-to-speech systems. This paper extends previous work on PbA in several directions. First, we have included full pattern matching between input letter string and dictionary entries, as well as including lexical stress in letter-to-phoneme conversion. Second, we have extended the method to phoneme-to-letter conversion. Third, and most important, we have experimented with multiple, different strategies for scoring the candidate pronunciations. Individual scores for each strategy are obtained on the basis of rank and either multiplied or summed to produce a final, overall score. Five strategies have been studied and results obtained from all 31 possible combinations. The two combination methods perform comparably, with the product rule only very marginally superior to the sum rule. Nonparametric statistical analysis reveals that performance improves as more strategies are included in the combination: this trend is very highly significant (p < 0:0005). Accordingly for letter-to-phoneme conversion, best results are obtained when all five strategies are combined: word accuracy is raised to 65.5 relative to 61.7 for our best previous result and 63.0 for the best-performing single strategy. These improvements are very highly significant (p 0 and p < 0:00011 respectively). Similar results were found for phoneme-to-letter and letter-to-stress conversion, although the former was an easier problem for PbA than letter-to-phoneme conversion and the latter was harder. The main sources of error for the multistrategy approach are very similar to those for the best single strategy, and mostly involve vowel letters and phonemes.


Computer Speech & Language | 1999

Evaluating the pronunciation component of text-to-speech systems for English: a performance comparison of different approaches

Robert I. Damper; Yannick Marchand; Martin J. Adamson; Kjell Gustafson

The automatic derivation of word pronunciations from input text is a central task for any text-to-speech system. For general English text at least, this is often thought to be a solved problem, with manually-derived linguistic rules assumed capable of handling “novel” words missing from the system dictionary. Data-driven methods, based on machine learning of the regularities implicit in a large pronouncing dictionary, have received considerable attention recently but are generally thought to perform less well. However, these tentative beliefs are at best uncertain without powerful methods for comparing text-to-phoneme subsystems. This paper contributes to the development of such methods by comparing the performance of four representative approaches to automatic phonemization on the same test dictionary. As well as rule-based approaches, three data-driven techniques are evaluated: pronunciation by analogy (PbA), NETspeak and IB1-IG (a modified k-nearest neighbour method). Issues involved in comparative evaluation are detailed and elucidated. The data-driven techniques outperform rules in accuracy of letter-to-phoneme translation by a very significant margin but require aligned text-phoneme training data and are slower. Best translation results are obtained with PbA at approximately 72% words correct on a resonably large pronouncing dictionary, compared with something like 26% words correct for the rules, indicating that automatic pronunciation of text is not a solved problem. c 1999 Academic Press


Clinical Neurophysiology | 2005

Assessment of working memory abilities using an event-related brain potential (ERP)-compatible digit span backward task ☆

Celeste D. Lefebvre; Yannick Marchand; Gail A. Eskes; John F. Connolly

OBJECTIVE This study investigated the effectiveness of an ERP-compatible Digit Span Backward (ERP-DB) task to determine working memory abilities in healthy participants. METHODS Participants were administered both the standard digit span backward and ERP-DB tasks. The ERP-DB task was divided into two sections, consisting of 2, 4, 6 and 8 (Group 1) and 3, 5, and 7 (Group 2) set sizes. A set of digits was aurally presented, followed by a second set that either corresponded to the reverse order of the first set (correct condition) or had one digit in the sequence replaced by an incorrect digit (incorrect condition). RESULTS Two posterior positive components were found to distinguish the two conditions; an earlier positive component (P200/P300) was elicited in the correct condition, whereas a comparatively robust and prolonged positive slow wave (PSW) was elicited in the incorrect condition. Furthermore, the PSW and the difference in PSW amplitude between incorrect and correct conditions (dPSW) dissipated as working memory load increased and were related to working memory capacity. CONCLUSIONS The PSW, dPSW and P200/P300 components were found to be associated with working memory abilities and may have the potential to act as neurophysiological markers for the assessment of working memory capacity. SIGNIFICANCE This research lends support for the utility of the ERP-DB task as a means of assessing working memory abilities, which may have implications for testing patients with expressive communication impairments.


Clinical Neurophysiology | 2002

Linking neurophysiological and neuropsychological measures for aphasia assessment

Yannick Marchand; Ryan C.N. D'Arcy; John F. Connolly

OBJECTIVES The objective of this study was to find the event-related brain potential (ERP) waveform features and parameters that maximize the correlation between the ERP components and behavioral performance on a neuropsychological test of language comprehension (PPVT-R) in order to develop an electrophysiological diagnostic technique that can be used in the assessment of aphasic patients. METHODS ERPs were recorded during a computerized version of the Peabody Picture Vocabulary Test-Revised (PPVT-R, Form M). In the computerized version, a picture is presented followed by a congruent or incongruent spoken word. A derived measure was calculated from the ERP differentiation between congruent and incongruent words. The traditional PPVT-R (Form L) was also administered for comparison purposes. The participants included 10 left-sided stroke patients. RESULTS The N400 was the primary component elicited to incongruent spoken words. Following optimization procedures, a statistical correlation (Pearson r=0.86) was found between the derived N400 measures and the neuropsychological test scores. Examination of the scatter plot confirmed that the relationship was linear. The derived N400 measure was defined primarily as the mean of the t-scores obtained from the incongruent and congruent waveform comparison, within the temporal interval that encompasses the N400. CONCLUSIONS This novel quantification technique links ERPs with neuropsychological data at an unprecedented level. Given the high correlation, a regression line could reasonably be used to estimate a patients language ability using only ERPs. However, before these findings can be accepted fully, these results need to be replicated in larger samples and across other paradigms.


Clinical Neurophysiology | 2003

Electrophysiological assessment of language function following stroke.

Ryan C.N. D'Arcy; Yannick Marchand; Gail A. Eskes; Edmund R. Harrison; Stephen Phillips; Alma Major; John F. Connolly

OBJECTIVE Event-related brain potentials (ERPs) were used to assess language function after stroke and demonstrate that it is possible to adapt neuropsychological tests to evaluate neurocognitive function using ERPs. Prior ERP assessment work has focused on language in both healthy individuals and case studies of aphasic neurotrauma patients. The objective of the current study was to evaluate left-hemisphere stroke patients who had varying degrees of receptive language impairment. It was hypothesized that ERPs would assess receptive language function accurately and correlate highly with the neuropsychological data. METHODS Data were collected from 10 left-hemisphere stroke patients; all were undergoing rehabilitation at the time of testing. Each patient received a battery of neuropsychological tests including the Peabody Picture Vocabulary Test-Revised (PPVT-R; Minnesota: American Guidance Service, 1981). ERPs were recorded during a computerized PPVT-R, in which pictures are presented followed by digitized spoken words that are either congruent or incongruent with the pictures. RESULTS AND CONCLUSION Incongruent spoken words within an individuals vocabulary level elicited well-known ERP components. One of the components (the N400) could be utilized as a marker of intact semantic processing. The ERP results were subsequently quantified and N400 derivative scores correlated highly with the neuropsychological findings. The results provided a clear demonstration of the efficacy of ERP-based assessment in a neurological patient group. SIGNIFICANCE Language function in stroke patients can be evaluated, independent of behavior, using electrophysiological measures that correlate highly with traditional neuropsychological test scores.


Natural Language Engineering | 2007

Can syllabification improve pronunciation by analogy of English

Yannick Marchand; Robert I. Damper

In spite of difficulty in defining the syllable unequivocally, and controversy over its role in theories of spoken and written language processing, the syllable is a potentially useful unit in several practical tasks which arise in computational linguistics and speech technology. For instance, syllable structure might embody valuable information for building word models in automatic speech recognition, and concatenative speech synthesis might use syllables or demisyllables as basic units. In this paper, we first present an algorithm for determining syllable boundaries in the orthographic form of unknown words that works by analogical reasoning from a database or corpus of known syllabifications. We call this syllabification by analogy (SbA). It is similarly motivated to our existing pronunciation by analogy (PbA) which predicts pronunciations for unknown words (specified by their spellings) by inference from a dictionary of known word spellings and corresponding pronunciations. We show that including perfect (according to the corpus) syllable boundary information in the orthographic input can dramatically improve the performance of pronunciation by analogy of English words, but such information would not be available to a practical system. So we next investigate combining automatically-inferred syllabification and pronunciation in two different ways: the series model in which syllabification is followed sequentially by pronunciation generation; and the parallel model in which syllabification and pronunciation are simultaneously inferred. Unfortunately, neither improves performance over PbA without syllabification. Possible reasons for this failure are explored via an analysis of syllabification and pronunciation errors.


International Journal of Speech Technology | 2005

Aligning Text and Phonemes for Speech Technology Applications Using an EM-Like Algorithm

Robert I. Damper; Yannick Marchand; J.-D. S. Marsters; Alex I. Bazin

A common requirement in speech technology is to align two different symbolic representations of the same linguistic ‘message’. For instance, we often need to align letters of words listed in a dictionary with the corresponding phonemes specifying their pronunciation. As dictionaries become ever bigger, manual alignment becomes less and less tenable yet automatic alignment is a hard problem for a language like English. In this paper, we describe the use of a form of the expectation-maximization (EM) algorithm to learn alignments of English text and phonemes, starting from a variety of initializations. We use the British English Example Pronunciation (BEEP) dictionary of almost 200,000 words in this work. The quality of alignment is difficult to determine quantitatively since no ‘gold standard’ correct alignment exists. We evaluate the success of our algorithm indirectly from the performance of a pronunciation by analogy system using the aligned dictionary data as a knowledge base for inferring pronunciations. We find excellent performance—the best so far reported in the literature. There is very little dependence on the start point for alignment, indicating that the EM search space is strongly convex. Since the aligned BEEP dictionary is a potentially valuable resource, it is made freely available for research use.


Language and Speech | 2009

Automatic Syllabification in English: A Comparison of Different Algorithms

Yannick Marchand; Connie R. Adsett; Robert I. Damper

Automatic syllabification of words is challenging, not least because the syllable is not easy to define precisely. Consequently, no accepted standard algorithm for automatic syllabification exists. There are two broad approaches: rule-based and data-driven. The rule-based method effectively embodies some theoretical position regarding the syllable, whereas the data-driven paradigm tries to infer “new” syllabifications from examples assumed to be correctly syllabified already. This article compares the performance of several variants of the two basic approaches. Given the problems of definition, it is difficult to determine a correct syllabification in all cases and so to establish the quality of the “gold standard” corpus used either to evaluate quantitatively the output of an automatic algorithm or as the example-set on which data-driven methods crucially depend. Thus, we look for consensus in the entries in multiple lexical databases of pre-syllabified words. In this work, we have used two independent lexicons, and extracted from them the same 18,016 words with their corresponding (possibly different) syllabifications. We have also created a third lexicon corresponding to the 13,594 words that share the same syllabifications in these two sources. As well as two rule-based approaches (Hammonds and Fishers implementation of Kahns), three data-driven techniques are evaluated: a look-up procedure, an exemplar-based generalization technique, and syllabification by analogy (SbA). The results on the three databases show consistent and robust patterns. First, the data-driven techniques outperform the rule-based systems in word and juncture accuracies by a very significant margin but require training data and are slower. Second, syllabification in the pronunciation domain is easier than in the spelling domain. Finally, best results are consistently obtained with SbA.


Information Fusion | 2006

Information fusion approaches to the automatic pronunciation of print by analogy

Robert I. Damper; Yannick Marchand

Automatic pronunciation of words from their spelling alone is a hard computational problem, especially for languages like English and French where there is only a partially consistent mapping from letters to sound. Currently, the best known approach uses an inferential process of analogy with other words listed in a dictionary of spellings and corresponding pronunciations. However, the process produces multiple candidate pronunciations and little or no theory exists to guide the choice among them. Rather than committing to one specific heuristic scoring method, it may be preferable to use multiple strategies (i.e., soft experts) and then employ information fusion techniques to combine them to give a final result. In this paper, we compare four different fusion schemes, using three different dictionaries (with different codings for specifying the pronunciations) as the knowledge base for analogical reasoning. The four schemes are: fusion of raw scores; rank fusion using Borda counting; rank fusion using non-uniform values; and rank fusion using non-uniform values weighted by a measure of prior performance of the experts. All possible combinations of five different expert strategies are studied. Although all four fusion schemes outperformed the single best strategy, results show clear superiority of rank fusion over the other methods.


Journal of Clinical and Experimental Neuropsychology | 2006

Event-related Brain Potentials as a Measure of Performance on WISC-III and WAIS-R NI Similarities Sub-tests

John F. Connolly; Yannick Marchand; Ryan C.N. D'Arcy

Communication difficulties are a common consequence of brain injury and neuropathological processes. Equally common is the inability to assess intellectual functioning in many communication-impaired populations because the aphasia and physical disability of such patients prevents measurable performance on traditionally administered tests of mental functioning. This study demonstrates that concept formation abilities can be assessed using event-related brain potentials. It also provides further evidence for the efficacy of this innovative assessment method in which standardized and validated psychometric tests are formatted for computer presentation and simultaneous recording of neural activity, which serves as the response measure in lieu of verbal/behavioral responses.

Collaboration


Dive into the Yannick Marchand's collaboration.

Top Co-Authors

Avatar
Top Co-Authors

Avatar
Top Co-Authors

Avatar
Top Co-Authors

Avatar
Top Co-Authors

Avatar
Top Co-Authors

Avatar
Top Co-Authors

Avatar
Top Co-Authors

Avatar
Top Co-Authors

Avatar

Kjell Gustafson

Royal Institute of Technology

View shared research outputs
Top Co-Authors

Avatar
Researchain Logo
Decentralizing Knowledge