This work outlines the problems encountered in modeling pronunciation for automatic speech recognition (ASR) of spontaneous (American) English speech. We detail some of the phonetic phenomena within the Switchboard corpus that make the recognition of this speaking style difficult. Phonetic transcribers found that feature spreading and cue trading made identification of phonetic segmental bounda...

Mispronunciation is still among the frailties of language learners the world over. The present study reports an all-out investigation into pronunciation training for the Persian-speaking EFL learners at the intermediate level of language proficiency. More specifically, the present study investigated (a) the underlying causes of mispronunciation at the level of segmental features, (b) the crucia...

A detailed description of the discourse context of a word can be used for predicting word pronunciation in discourse context and also enables studies of the interplay between various types of information on e.g. phone-level pronunciation. The work presented in this paper is aimed at modelling systematic variation in the phone-level realisation of words inherent to a language variety. A data-dri...

Mispronunciation is still among the frailties of language learners the world over. The present study reports an all-out investigation into pronunciation training for the Persian-speaking EFL learners at the intermediate level of language proficiency. More specifically, the present study investigated (a) the underlying causes of mispronunciation at the level of segmental features, (b) the crucia...

Consonants in word onsets are, in English and Dutch, more frequently misspoken than consonants in other positions, and also more frequently than expected from the relative numbers of onset consonants and other consonants. We argue here that relative numbers of segments in specific positions in the word is not a valid predictor of relative frequencies of segmental speech errors. A more valid pre...

This paper proposes a paradigm where commonly made segmental pronunciation errors are modeled as pair-wise confusions between two or more phonemes in the language that is being learnt. The method uses an ensemble of support vector machine classifiers with time varying Mel frequency cepstral features to distinguish between several pairs of phonemes. These classifiers are then applied to classify...

This study examined how phonological information is represented and retrieved, using analysis of naming errors in anomia. We focused on questions that relate to metrical and segmental information, the types of information each of them contains, and whether they are organised in parallel or serially. Nine Hebrew-speaking individuals with anomia due to phonological output deficit named 200 pictur...

When pronouncing novel/unknown words, readers often use prior experience with similar, neighbor words. Comparison to neighbors can be helpful for unknown or novel words (wug is like pug), but it can also lead to errors (pint is not like mint). We investigate whether pronunciation can be affected by top-down influences, specifically the perceived taboo-ness of a known neighbor. While orthographi...

In leading morpho-phonological theories and state-of-the-art text-to-speech systems it is assumed that word pronunciation cannot be learned or performed without in-between analyses at several abstraction levels (e.g., morphological, graphemic, phonemic, syllabic, and stress levels). We challenge this assumption for the case of English word pronunciation. Using igtree, an inductive-learning deci...