{U} sing automatic speech recognition for phonological purposes:{S} tudy of {V} owel {L} enght in {P} unu ({B} antu {B} 40) (original) (raw)
Related papers
A phonetic study of Nanai vowels: Using automated post-transcriptional processing techniques
ALTAI HAKPO, 2016
KO Seongyeon (CUNY Queens College) YUN, Jiwon, Hijo KANG & Seongyeon KO. 2016. A phonetic study of Nanai vowels: Using automated post-transcriptional processing techniques. Altai Hakpo 26. 29-44. The Altaic Society of Korea. Recent years have witnessed a remarkable growth in Altaic field linguistics, with recognizable achievements especially in language description and archiving. However, relatively less effort has been made to analyze the collected materials from the general phonetic perspectives. In this paper, the authors demonstrate how the archived language data can be analyzed using automated tools for speech annotation and phonetic analysis and thus with drastically reduced time and effort. Using the published IPA-based transcriptions of the lexical items (Ko & Yurn 2011), we ran Prosodylabaligner (Gorman et al. 2011) for the automatic segmentation and labelling of the Nanai recordings collected by the Altaic Society of Korea. After this post-processing, the acoustic characteristics of vowels were automatically measured using scripts written for Praat (Boersma & Weenink 2014). The acoustic analysis of Nanai vowels shows that, by and large, the automated process produces relatively reliable results with its effectiveness evidenced in terms of time and effort. We believe that Altaic linguistics will be highly benefited from these practical techniques with the improvement of accuracy.
Automatic Phonetic Transcription: An Overview
Both in linguistics and in speech technology phonetic transcriptions (PTs) are often needed. Given the many drawbacks in making manual PTs, researchers have been looking for ways to obtain PTs automatically. In this paper an overview is presented of automatic phonetic transcription (APT). Several aspects of APT are discussed: evaluation, generation and usability. Evaluation is needed to determine the quality of APTs. Usually this is done by comparing the APTs with human reference transcriptions. Generating APTs can be done in several ways, e.g. by means of phone recognition or forced recognition. The quality of the generated APTs can be enhanced by optimizing the automatic speech recognition systems used to make the APTs. In spite of the current limitations of ASR technology, APTs already offer some important advantages for phonetic research. In this paper we explain how.
Machine Assisted Analysis of Vowel Length Contrasts in Wolof
Interspeech 2017
Growing digital archives and improving algorithms for automatic analysis of text and speech create new research opportunities for fundamental research in phonetics. Such empirical approaches allow statistical evaluation of a much larger set of hypothesis about phonetic variation and its conditioning factors (among them geographical / dialectal variants). This paper illustrates this vision and proposes to challenge automatic methods for the analysis of a not easily observable phenomenon: vowel length contrast. We focus on Wolof, an under-resourced language from Sub-Saharan Africa. In particular, we propose multiple features to make a fine evaluation of the degree of length contrast under different factors such as: read vs semispontaneous speech ; standard vs dialectal Wolof. Our measures made fully automatically on more than 20k vowel tokens show that our proposed features can highlight different degrees of contrast for each vowel considered. We notably show that contrast is weaker in semi-spontaneous speech and in a non standard semi-spontaneous dialect.
ArXiv, 2018
Corpus phonetics has become an increasingly popular method of research in linguistic analysis. With advances in speech technology and computational power, large scale processing of speech data has become a viable technique. This tutorial introduces the speech scientist and engineer to various automatic speech processing tools. These include acoustic model creation and forced alignment using the Kaldi Automatic Speech Recognition Toolkit (Povey et al., 2011), forced alignment using FAVE-align (Rosenfelder et al., 2014), the Montreal Forced Aligner (McAuliffe et al., 2017), and the Penn Phonetics Lab Forced Aligner (Yuan & Liberman, 2008), as well as stop consonant burst alignment using AutoVOT (Keshet et al., 2014). The tutorial provides a general overview of each program, step-by-step instructions for running the program, as well as several tips and tricks.
Phonetic data and phonological analyses
Stellenbosch Papers in Linguistics, 2012
This paper is basically concerned with the relationship between phonetic data and phonological analyses. I) It will be shown that phonological analyses based on unverified phonetic data tend to accommodate ad hoc, unmotivated, and even phonetically implausible phonological rules. On the other hand, it will be demonstrated that a phonological analysis (of the same phenomenon), based on verified phonetic data, accounts for these data in an acceptable, natural and credible manner. Examples will be taken from the phenomenon of labialization in Sesotho 2) to illustrate the point that it is absolutely necessary to make a clearcut distinction between "data" and "facts" in generative phonological descriptions. Attention will first be given to different types of phonetic data: impressionistic phonetic data of Tucker (1929) and of Kunene (i961) will be discussed, after which some experimental phonetic data on the phenomenon of labialization in Sesotho will be presented. A phonological analysis of Ponelis (1974) based on unverified phonetic data will then be examined, after which finally, a phonological analysis based on verified data will be considered.
Linguistic Phonetics: A Look into the Future
ICPhS, 2011
This paper exemplifies the need for collecting extensive experimental data on the languagedependent articulatory and acoustic characteristics of vowels and consonants. It reports data on constriction location for (alveolo)palatal consonants, and on the F2 frequency for allophonic realizations of different varieties of /l/, in several language groups.
The Handbook of Phonetic Sciences
This outstanding multi-volume series covers all the major subdisciplines within linguistics today and, when complete, will offer a comprehensive survey of linguistics as a whole.
Corpus Phonetics for Under-Documented Languages: A Vowel Harmony Example
2020
Corpus phonetics is enabling the comprehensive analysis of large digital speech collections. In this paper, we develop a corpus phonetics workflow that is flexible enough to be easily applied to under-documented languages. To test the capabilities of this workflow we choose a challenging vowel reduction and vowel harmony problem. In Kera (Chadic) it has been shown (Pearce, 2012), that not only is phonetic reduction linked to the phonetic duration of the vowel, but also that reduction is blocked in vowel harmony domains. We are able to replicate previously published experiments by Pearce that were originally completed using manual measurements. We expect that our corpus phonetics workflow will be of value to phonologists working on other languages.