This meta-analytic study explores the overall effectiveness of automatic speech recognition (ASR) on ESL/EFL student pronunciation performance. Data with 15 studies representing 38 effect sizes found from 2008 to 2021 were meta-analyzed. The findings of the meta-analysis indicated that ASR has a medium overall effect size (g = 0.69). Results from moderator analyses suggest that (1) ASR with explicit corrective feedback is largely effective, while ASR with indirect feedback (e.g. ASR dictation) is moderately effective; (2) ASR has a large effect on segmental pronunciation but a small effect on suprasegmental pronunciation; (3) medium to long treatment duration of ASR results in higher learning outcomes, but short duration offers no differential effect compared to a non-ASR condition; (4) practicing pronunciation with peers in an ASR condition produces a large effect, but the effect is small when practicing alone; (5) ASR is largely effective for adult (i.e. 18 years old and above) and intermediate English learners. Overall, ASR is a beneficial application and is recommended for assisting L2 student pronunciation development.

An increasing number of studies are exploring the benefits of automatic speech recognition (ASR)–based dictation programs for second language (L2) pronunciation learning (e.g. Chen, Inceoglu & Lim, 2020; Liakin, Cardoso & Liakina, 2015; McCrocklin, 2019), but how ASR recognizes accented speech and the nature of the feedback it provides to language learners is still largely under-researched. The current study explores whether the intelligibility of L2 speakers differs when assessed by native (L1) listeners versus ASR technology, and reports on the types of intelligibility issues encountered by the two groups. Twelve L1 listeners of English transcribed 48 isolated words targeting the /ɪ-i/ and /æ-ε/ contrasts and 24 short sentences that four Taiwanese intermediate learners of English had produced using Google’s ASR dictation system. Overall, the results revealed lower intelligibility scores for the word task (ASR: 40.81%, L1 listeners: 38.62%) than the sentence task (ASR: 75.52%, L1 listeners: 83.88%), and highlighted strong similarities in the error types – and their proportions – identified by ASR and the L1 listeners. However, despite similar recognition scores, correlations indicated that the ASR recognition of the L2 speakers’ oral productions mirrored the L1 listeners’ judgments of intelligibility in the word and sentence tasks for only one speaker, with significant positive correlations for one additional speaker in each task. This suggests that the extent to which ASR approaches L1 listeners at recognizing accented speech may depend on individual speakers and the type of oral speech.

The analysis of political texts from parliamentary speeches, party manifestos, social media, or press releases forms the basis of major and growing fields in political science, not least since advances in “text-as-data” methods have rendered the analysis of large text corpora straightforward. However, a lot of sources of political speech are not regularly transcribed, and their on-demand transcription by humans is prohibitively expensive for research purposes. This class includes political speech in certain legislatures, during political party conferences as well as television interviews and talk shows. We showcase how scholars can use automatic speech recognition systems to analyze such speech with quantitative text analysis models of the “bag-of-words” variety. To probe results for robustness to transcription error, we present an original “word error rate simulation” (WERSIM) procedure implemented in $R$. We demonstrate the potential of automatic speech recognition to address open questions in political science with two substantive applications and discuss its limitations and practical challenges.

This paper introduces a novel captioning method, partial and synchronized captioning (PSC), as a tool for developing second language (L2) listening skills. Unlike conventional full captioning, which provides the full text and allows comprehension of the material merely by reading, PSC promotes listening to the speech by presenting a selected subset of words, where each word is synched to its corresponding speech signal. In this method, word-level synchronization is realized by an automatic speech recognition (ASR) system, dedicated to the desired corpora. This feature allows the learners to become familiar with the correspondences between words and their utterances. Partialization is done by automatically selecting words or phrases likely to hinder listening comprehension. In this work we presume that the incidence of infrequent or specific words and fast delivery of speech are major barriers to listening comprehension. The word selection criteria are thus based on three factors: speech rate, word frequency and specificity. The thresholds for these features are adjusted to the proficiency level of the learners. The selected words are presented to aid listening comprehension while the remaining words are masked in order to keep learners listening to the audio. PSC was evaluated against no-captioning and full-captioning conditions using TED videos. The results indicate that PSC leads to the same level of comprehension as the full-captioning method while presenting less than 30% of the transcript. Furthermore, compared with the other methods, PSC can serve as an effective medium for decreasing dependence on captions and preparing learners to listen without any assistance.

The paper addresses a robust wavelet-based speech enhancement for automatic speech recognition in reverberant and noisy conditions. We propose a novel scheme in improving the speech, late reflection, and noise power estimates from the observed contaminated signal. The improved estimates are used to calculate the Wiener gain in filtering the late reflections and additive noise. In the proposed scheme, optimization of the wavelet family and its parameters is conducted using an acoustic model (AM). In the offline mode, the optimal wavelet family is selected separately for the speech, late reflections, and background noise based on the AM likelihood. Then, the parameters of the selected wavelet family are optimized specifically for each signal subspace. As a result we can use a wavelet sensitive to the speech, late reflection, and the additive noise, which can independently and accurately estimate these signals directly from an observed contaminated signal. For speech recognition, the most suitable wavelet is identified from the pre-stored wavelets, and wavelet-domain filtering is conducted to the noisy and reverberant speech signal. Experimental evaluations using real reverberant data demonstrate the effectiveness and robustness of the proposed method.

Oral communication ability has become increasingly important to many EFL students. Several commercial software programs based on automatic speech recognition (ASR) technologies are available but their prices are not affordable for many students. This paper will demonstrate how the Microsoft Speech Application Software Development Kit (SASDK), a free but powerful tool, can be used to develop an oral skills training website for EFL students. This ASR-based website offers six different types of online exercises which allow students to practise their oral skills and obtain immediate feedback on their performance. A group of 25 college students and a group of 35 pre-service English teachers were invited to use the website. Two surveys were conducted to investigate the students’ and the pre-service teachers’ perceptions of this site. The results indicated that most teachers and students enjoyed using this website, which they felt could help improve their English oral skills. They also pointed out that the main strength of the ASR-based learning system is that it offers several different types of exercises which can encourage learners to produce more output in a low-anxiety environment. The major limitations of the website are the insufficient feedback and the challenging standards one must meet in order to achieve a pass mark. These findings can be useful for teachers who are interested in using ASR in teaching and for CALL researchers who aim to develop better ASR-based systems for language learning.

Although the success of automatic speech recognition (ASR)-based Computer Assisted Pronunciation Training (CAPT) systems is increasing, little is known about the pedagogical effectiveness of these systems. This is particularly regrettable because ASR technology still suffers from limitations that may result in the provision of erroneous feedback, possibly leading to learning breakdowns. To study the effectiveness of ASR-based feedback for improving pronunciation, we developed and tested a CAPT system providing automatic feedback on Dutch phonemes that are problematic for adult learners of Dutch. Thirty immigrants who were studying Dutch were assigned to three groups using either the ASR-based CAPT system with automatic feedback, a CAPT system without feedback, or no CAPT system. Pronunciation quality was assessed for each participant before and after the training by human experts who evaluated overall segmental quality and the quality of the phonemes addressed in the training. The participants' impressions of the CAPT system used were also studied through anonymous questionnaires. The results on global segmental quality show that the group receiving ASR-based feedback made the largest mean improvement, but the groups' mean improvements did not differ significantly. The group receiving ASR-based feedback showed a significantly larger improvement than the no-feedback group in the segmental quality of the problematic phonemes targeted.