Accurate Speech Research Articles

A comparative study of machine learning-based Chinese dialect speech recognition

As the official language of China, Mandarin is used on all formal occasions. However, due to China's rich history and vast population, dialects have become the primary language spoken by many people. While not taught in school, dialects serve as unofficial languages in China and are passed down through word of mouth. In recent years, the use of dialects has significantly declined as people have adapted to using official Mandarin in all settings. However, among the elderly, dialects remain the primary language of communication. The great need to use dialects as the primary mode of communication has laid the foundation for the importance and high demand for dialect speech recognition. Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, research on this topic is further analysed and compared. The results suggest that there has been a certain degree of development and progress in dialect speech recognition technology. However, further research is needed to overcome the limitations of low-resource language speech recognition.

RETRACTED ARTICLE: A review on emotion recognition from dialect speech using feature optimization and classification techniques

RETRACTED ARTICLE: A review on emotion recognition from dialect speech using feature optimization and classification techniques

Assessment and analysis of accents in air traffic control speech: a fusion of deep learning and information theory.

Enhancing the generalization and reliability of speech recognition models in the field of air traffic control (ATC) is a challenging task. This is due to the limited storage, difficulty in acquisition, and high labeling costs of ATC speech data, which may result in data sample bias and class imbalance, leading to uncertainty and inaccuracy in speech recognition results. This study investigates a method for assessing the quality of ATC speech based on accents. Different combinations of data quality categories are selected according to the requirements of different model application scenarios to address the aforementioned issues effectively. The impact of accents on the performance of speech recognition models is analyzed, and a fusion feature phoneme recognition model based on prior text information is constructed to identify phonemes of speech uttered by speakers. This model includes an audio encoding module, a prior text encoding module, a feature fusion module, and fully connected layers. The model takes speech and its corresponding prior text as input and outputs a predicted phoneme sequence of the speech. The model recognizes accented speech as phonemes that do not match the transcribed phoneme sequence of the actual speech text and quantitatively evaluates the accents in ATC communication by calculating the differences between the recognized phoneme sequence and the transcribed phoneme sequence of the actual speech text. Additionally, different levels of accents are input into different types of speech recognition models to analyze and compare the recognition accuracy of the models. Experimental results show that, under the same experimental conditions, the highest impact of different levels of accents on speech recognition accuracy in ATC communication is 26.37%. This further demonstrates that accents affect the accuracy of speech recognition models in ATC communication and can be considered as one of the metrics for evaluating the quality of ATC speech.

The Accent Atlas: Effects of long-term exposure to nonnative accents on adaptive speech perception

As mobility increases and virtual interactions grow, quickly adapting to diverse accents is essential for effective communication. While lab-based accent training promotes rapid adaptation to unfamiliar accents, whether prolonged environmental exposure to linguistic diversity yields similar benefits is unknown. We conducted a large-scale perceptual experiment involving 600 + participants across 15 U.S. states, utilizing a cross-modal word matching task to measure speech recognition in both non-native accented English and native English with background noise. Participants were recruited from linguistically diverse (LD) or linguistically homogeneous (LH) states. Linguistic diversity was defined based on the prevalence of non-English speaking residents (U.S. census data). Our findings were twofold: firstly, consistent with prior studies, only exposure to non-native accented speech—but not exposure to speech in noise—improved recognition of novel talkers with the same accent; secondly, individuals from the LD states outperformed those from LH areas in initial accented speech perception. However, by the end of the experiment, benefits from in-lab accent exposure did not differ significantly between the two groups, suggesting that short-term training could mitigate long-term environmental influences. These results illuminate the interplay between daily linguistic environments and accent adaptation, enhancing our understanding of speech perception's adaptivity in a multilingual context.

Investigation of machine-learning-based stimuli for the remediation of children’s speech errors

Children with speech sound disorders (SSDs) demonstrate difficulty producing phonemes correctly and may exhibit poor speech perception compared to their age-matched peers; however, group differences in speech perception skills remain largely unexplained. Developmental models of speech production posit that children’s ability to discriminate correct and incorrect productions in their own speech may be critical for developing accurate speech production. Historically, when children regularly mispronounce a phoneme, it has been essentially impossible to assess whether they perceive correct versus errors productions in their own speech, thus creating a clinical conundrum. How can we assess a child’s ability to perceive the accuracy of their own phoneme production when they cannot produce a correct production? Recent technological developments allow for acoustic alteration of children’s speech that digitally corrects speech sound errors while preserving natural characteristics of the child’s voice. This machine-learning-based stimuli may then be used as training and feedback tokens for remediation when treating children with speech sound disorders. Such acoustic alteration is possible within an accessible, user-friendly environment that is clinically feasible for speech-language pathologists with little acoustic training. Thus, the purpose of this study is to evaluate the acoustic and perceptual accuracy of acoustically-altered child-speech compared to natural speech tokens.

SR-TTS: a rhyme-based end-to-end speech synthesis system.

Deep learning has significantly advanced text-to-speech (TTS) systems. These neural network-based systems have enhanced speech synthesis quality and are increasingly vital in applications like human-computer interaction. However, conventional TTS models still face challenges, as the synthesized speeches often lack naturalness and expressiveness. Additionally, the slow inference speed, reflecting low efficiency, contributes to the reduced voice quality. This paper introduces SynthRhythm-TTS (SR-TTS), an optimized Transformer-based structure designed to enhance synthesized speech. SR-TTS not only improves phonological quality and naturalness but also accelerates the speech generation process, thereby increasing inference efficiency. SR-TTS contains an encoder, a rhythm coordinator, and a decoder. In particular, a pre-duration predictor within the cadence coordinator and a self-attention-based feature predictor work together to enhance the naturalness and articulatory accuracy of speech. In addition, the introduction of causal convolution enhances the consistency of the time series. The cross-linguistic capability of SR-TTS is validated by training it on both English and Chinese corpora. Human evaluation shows that SR-TTS outperforms existing techniques in terms of speech quality and naturalness of expression. This technology is particularly suitable for applications that require high-quality natural speech, such as intelligent assistants, speech synthesized podcasts, and human-computer interaction.

Analysis of forced aligner performance on L2 English speech

There is growing interest in how speech technologies perform on L2 speech. Largely omitted from this discussion are tools used in the early data processing steps, such as forced aligners, that can introduce errors and biases. This study adds to the conversation and tests how well a model pre-trained for the alignment of L1 American English speech performs on L2 English speech. We test and discuss the impact of language variety, demographic factors, and segment type on the performance of the forced aligner. We also examine systematic errors encountered.Forty-five speakers representing nine L2 varieties were selected from the Speech Accent Archive and force aligned using the Montreal Forced Aligner. The phoneme-level boundary placements were manually corrected in order to assess differences between the automatic and manual alignments. Results show marked variation in the performance across language groups and segment types for the two metrics used to assess accuracy: Onset Boundary Displacement, a distance metric between the automatic and manual boundary placements, and Overlap Rate, which indicates to what extent the automatically aligned segment overlaps with the manually aligned segment. The highest accuracy on both measures was obtained for German and French, and lowest accuracy for Russian. The aligner's performance on all varieties was comparable to that on conversational American English and non-standard varieties of English. Furthermore, the percentage of boundary placements within 10 and 20 ms of the corrected boundary was similar to that observed between transcribers. Apart from errors due to variety mismatch, most issues encountered in the alignment were due to issues not exclusive to L2 speech such as inaccurate orthographic transcriptions, hesitations, specific voice qualities, and background noise.The results of this study can inform the use of automatic aligners on L2 English speech and provide a baseline of potential errors and information to help the development of more robust alignment tools for further development of automatic systems using L2 English.

HoloGCS: mixed reality-based ground control station for unmanned aerial vehicle

Human–robot interaction (HRI), which studies the interaction between robots and humans, appears as a promising research idea for the future of smart factories. In this study, HoloLens as ground control station (HoloGCS) is implemented, and its performance is discussed. HoloGCS is a mixed reality-based system for controlling and monitoring unmanned aerial vehicles (UAV). The system incorporates HRI through speech commands and video streaming, enabling UAV teleoperation. HoloGCS provides a user interface that allows operators to monitor and control the UAV easily. To demonstrate the feasibility of the proposed systems, a user case study (user testing and SUS-based questionnaire) was performed to gather qualitative results. In addition, throughput, RTT, latency, and speech accuracy were also gathered and analyzed to evaluate quantitative results.

Outcome Prediction of Speech Perception in Quiet and in Noise for Cochlear Implant Candidates Based on Pre-Operative Measures.

(1) Background: The fitting of cochlear implants (CI) is an established treatment, even in cases with considerable residual hearing but insufficient speech perception. The aim of this study was to evaluate a prediction model for speech in quiet and to provide reference data and a predictive model for postoperative speech perception in noise (SPiN) after CI provision. (2) Methods: CI candidates with substantial residual hearing (either in hearing threshold or in word recognition scores) were included in a retrospective analysis (n = 87). Speech perception scores in quiet 12 months post-surgery were compared with the predicted scores. A generalized linear model was fitted to speech reception thresholds (SRTs) after CI fitting to identify predictive variables for SPiN. (3) Results: About two-thirds of the recipients achieved the expected outcome in quiet or were better than expected. The mean absolute error of the prediction was 13.5 percentage points. Age at implantation was the only predictive factor for SPiN showing a significant correlation (r = 0.354; p = 0.007). (4) Conclusions: Outcome prediction accuracy for speech in quiet was comparable to previous studies. For CI recipients in the included study population, the SPiN outcome could be predicted only based on the factor age.

Discrimination of Second Language Vowel Contrasts and the Role of Phonological Short-Term Memory and Nonverbal Intelligence

Although extensive research has focused on the perceptual abilities of second language (L2) learners, a significant gap persists in understanding how cognitive functions like phonological short-term memory (PSTM) and nonverbal intelligence (IQ) impact L2 speech perception. This study sets out to investigate the discrimination of L2 English monophthongal vowel contrasts and to assess the effect of PSTM and nonverbal IQ on L2 speech perception. The participants consisted of adult monolingually-raised Greek speakers, who completed an AX discrimination test, a digit span test, and a nonverbal intelligence test. A control group of English speakers also completed the AX test. Data were analyzed using Bayesian regression models. The results revealed that Greek speakers exhibited below chance discrimination for the majority of L2 vowel contrasts, consistently underperforming in comparison to the control group. Intriguingly, the study did not provide substantial evidence in favor of more accurate discrimination of L2 contrasts by Greek participants with high PSTM compared to those with low PSTM. However, the study yielded compelling evidence indicating that Greek participants with higher IQ demonstrated superior accuracy in discriminating most L2 contrasts compared to their lower IQ counterparts. The limited influence of PSTM on speech perception suggests the need for further exploration, considering the potential impact of test methodologies and the intricate interplay of other confounding factors. Furthermore, the study uncovers a noteworthy relationship between nonverbal IQ and L2 speech perception, likely linked with the association of high IQ with enhanced attentional capacities, information processing abilities, and learning skills—all of which are pivotal for accurate speech perception.

Chinese dialect speech recognition: a comprehensive survey

As a multi-ethnic country with a large population, China is endowed with diverse dialects, which brings considerable challenges to speech recognition work. In fact, due to geographical location, population migration, and other factors, the research progress and practical application of Chinese dialect speech recognition are currently at different stages. Therefore, exploring the significant regional heterogeneities in specific recognition approaches and effects, dialect corpus, and other resources is of vital importance for Chinese speech recognition work. Based on this, we first start with the regional classification of dialects and analyze the pivotal acoustic characteristics of dialects, including specific vowels and tones patterns. Secondly, we comprehensively summarize the existing dialect phonetic corpus in China, which is of some assistance in exploring the general construction methods of dialect phonetic corpus. Moreover, we expound on the general process of dialect recognition. Several critical dialect recognition approaches are summarized and introduced in detail, especially the hybrid method of Artificial Neural Network (ANN) combined with the Hidden Markov Model(HMM), as well as the End-to-End (E2E). Thirdly, through the in-depth comparison of their principles, merits, disadvantages, and recognition performance for different dialects, the development trends and challenges in dialect recognition in the future are pointed out. Finally, some application examples of dialect speech recognition are collected and discussed.

Transfer learning methods for low-resource speech accent recognition: A case study on Vietnamese language

Speech accent recognition (SAR) plays a crucial role in enhancing communication between customers and service providers, enabling personalized interactions based on geographical, birthplace, and cultural cues derived from accents. However, current approaches predominantly rely on training SAR models from scratch, overlooking the potential of transfer learning from other speech processing tasks, despite the relatively small size of accent datasets. This paper represents the first comprehensive investigation into the effectiveness of transfer learning methods derived from a diverse array of data-rich speech processing tasks for SAR. Through experiments on a practical Vietnamese telephone dataset provided by Viettel, the largest telecommunications provider in Southeast Asia, our study reveals that our best-performing model outperforms previous state-of-the-art SAR models by 46.7% in terms of relative accuracy.

The impact of non-native English speakers’ phonological and prosodic features on automatic speech recognition accuracy

The present study examines the impact of Arab speakers’ phonological and prosodic features on the accuracy of automatic speech recognition (ASR) of non-native English speech. The authors first investigated the perceptions of 30 Egyptian ESL teachers and 70 Egyptian university students towards the L1 (Arabic)-based errors affecting intelligibility and then carried out a data analysis of the ASR of the students’ English speech to find out whether the errors investigated resulted in intelligibility breakdowns in an ASR setting. In terms of the phonological features of non-native speech, the results showed that the teachers gave more weight to pronunciation features of accented speech that did not actually hinder recognition, that the students were mostly oblivious to the L2 errors they made and their impact on intelligibility, and that L2 errors which were not perceived as serious by both teachers and students had negative impacts on ASR accuracy levels. In regard to the prosodic features of non-native speech, it was found that lower speech rates resulted in more accurate speech recognition levels, higher speech intensity led to less deletion errors, and voice pitch did not seem to have any impact on ASR accuracy levels. The study, accordingly, recommends training ASR systems with more non-native data to increase their accuracy levels as well as paying more attention to remedying non-native speakers’ L1-based errors that are more likely to impact non-native automatic speech recognition.

Biodegradable, Bifunctional Electro-acoustic Transducers Based on Cellular Polylactic Acid Ferroelectrets for Sustainable Flexible Electronics.

Nowadays, humans rely increasingly on smart electronics to address grand challenges and to improve life conditions in the era of digitalization and big data. However, electronics often have a limited lifespan, and they may bring electronic waste problems after their service. To mitigate this problem, environmentally sustainable methods of electronic device production and disposal are highly recommended, where advanced functional materials should be redesigned with improved sensing performance over the entire operational life while also being naturally degradable at the end. Herein, a biodegradable and flexible bifunctional electroacoustic transducer was fabricated with the utilization of cellular polylactic acid (PLA) ferroelectret films, possessing a small acoustic impedance of ∼0.02 MRayl which is quite close to that of air and a high figure of merit (FOM: d33·g33) of ∼11 GPa-1. Such devices have a prominent signal-to-noise ratio (SNR) of ∼23.5 dB @1 kHz and can work either as a microphone by direct piezoelectric effect or a loudspeaker by reverse piezoelectric effect in air medium. When used as a microphone, the flexible device exhibits a prominent receiving sensitivity up to 4.2 mV/Pa (∼-47.5 dB/ref. 1 V/Pa) at 1 kHz. When served as a loudspeaker, it is capable of yielding high sound pressure levels (SPLs) ranging from 60 to 103 dB (ref. 20 μPa) in a broad frequency range of 1-80 kHz with an active area of 3.14 cm2. Additionally, the electrical response curve of the device is very flat in a wide frequency range from 300 to 3000 Hz. With the high-performance acoustic-electric conversion capacity, the PLA ferroelectret-based flexible and filmlike electroacoustic transducer was used to realize accurate speech recognition and control, providing a strong impetus for its advanced and eco-friendly applications in the era of the internet of things (IoT) and artificial intelligence.

Subject-independent meta-learning framework towards optimal training of EEG-based classifiers

Advances in deep learning have shown great promise towards the application of performing high-accuracy Electroencephalography (EEG) signal classification in a variety of tasks. However, many EEG-based datasets are often plagued by the issue of high inter-subject signal variability. Robust deep learning models are notoriously difficult to train under such scenarios, often leading to subpar or widely varying performance across subjects under the leave-one-subject-out paradigm. Recently, the model agnostic meta-learning framework was introduced as a way to increase the model’s ability to generalize towards new tasks. While the original framework focused on task-based meta-learning, this research aims to show that the meta-learning methodology can be modified towards subject-based signal classification while maintaining the same task objectives and achieve state-of-the-art performance. Namely, we propose the novel implementation of a few/zero-shot subject-independent meta-learning framework towards multi-class inner speech and binary class motor imagery classification. Compared to current subject-adaptive methods which utilize large number of labels from the target, the proposed framework shows its effectiveness in training zero-calibration and few-shot models for subject-independent EEG classification. The proposed few/zero-shot subject-independent meta-learning mechanism performs well on both small and large datasets and achieves robust, generalized performance across subjects. The results obtained shows a significant improvement over the current state-of-the-art, with the binary class motor imagery achieving 88.70% and the accuracy of multi-class inner speech achieving an average of 31.15%. Codes will be made available to public upon publication.

Patient with Foreign Accent Syndrome in Post Infarct Thrombotic Stroke: A Case Report

Background: Stroke is the second leading cause of death in the world and the leading cause of disability worldwide. One of the disabilities is language disorder. Foreign Accent Syndrome (FAS) is a clinical condition associated with a noticeable change of accent, affecting verbal communication and social interaction skills. By identifying this case earlier, we hope the patient’s daily living is less disrupted. Case: A 54-year-old woman came with complaints of weakness and tingling in the right limbs accompanied by a change in language accent. Weakness in the right limbs was felt since 4 years ago, then 1 year later the patient began to experience a change in the speech accent, from a Javanese accent to a Madurese accent, where previously the patient had never lived or studied Madurese. The results of the neurobehavior examination showed the patient had disturbances in the language and memory domain. MRI revealed cerebral infarction in the bilateral hippocampi, small vessel ischemic in the bilateral frontal lobes, and corona radiata. Discussion: The accent change, known as FAS experienced by the patients, is thought to be due to neurological causes, which was stroke. In a study using a lesion network mapping approach, it was concluded that the lesion causing accent disorder was located within a single network in the bilateral frontal lobe. In line with the existing theory, the MRI results in this patient showed ischemia in the blood vessels of the bilateral frontal lobes.

Sacral phraseology in dialectal speech

Sacral phraseology in dialectal speech

Իրանական լեզուների հայերեն անունները. ժամանակագրությունը, գործածությունը, բառարանային մշակումը և կանոնարկման խնդիրները / Armenian Names of Iranian Languages: Chronology, Usage, Lexicographical Processing, and Normative Issues

In the fifth century, the Armenian language used the names of two Iranian languages: պարսկերէն (Persian) and պարթեւարէն (Parthian). In the seventh century, the linguonym պարսիկ լեզու (Persian) was recorded, while in Middle Armenian it was ֆարսի (Farsi). The formation of the remaining linguonyms refers to modern times. In the Armenian language, linguonyms of living and dead Iranian languages are in some cases names of dialects. They are usually formed using the suffix –երեն (-eren). The exceptions are the ancient Armenian պարթեւարէն (Parthian language) and the variable սոգդիարեն (Sogdian language), formed in the twentieth century, using the suffix -արէն (–aren). The Armenian names of Iranian languages have the following options: a) phonetic-orthographic, b) word-formative, c) lexical, d) structural (word and phrase). In the Armenian scientific literature, in particular in the linguistic works on Iranian studies, the variability of Armenian and borrowed linguonyms is noted. It is necessary to minimize the variability of Armenian names of Iranian languages and unify linguonyms, which will contribute to the correctness and accuracy of scientific speech. In monolingual and bilingual dictionaries of the Old Armenian and New Armenian languages, the names of two dozen Iranian languages with their various variants became headwords. The same number of linguonyms, including the names of dialects, have not yet been recorded in dictionaries. Among the Armenian names of Iranian languages, the most commonly used linguonyms are պարսկերեն (Persian language), տաջիկերեն (Tajik language) and քրդերեն (Kurdish language). A number of linguonyms are found only in scientific texts, some are used in isolated cases.

From first encounters to longitudinal exposure: a repeated exposure-test paradigm for monitoring speech adaptation.

Perceptual difficulty with an unfamiliar accent can dissipate within short time scales (e.g., within minutes), reflecting rapid adaptation effects. At the same time, long-term familiarity with an accent is also known to yield stable perceptual benefits. However, whether the long-term effects reflect sustained, cumulative progression from shorter-term adaptation remains unknown. To fill this gap, we developed a web-based, repeated exposure-test paradigm. In this paradigm, short test blocks alternate with exposure blocks, and this exposure-test sequence is repeated multiple times. This design allows for the testing of adaptive speech perception both (a) within the first moments of encountering an unfamiliar accent and (b) over longer time scales such as days and weeks. In addition, we used a Bayesian ideal observer approach to select natural speech stimuli that increase the statistical power to detect adaptation. The current report presents results from a first application of this paradigm, investigating changes in the recognition accuracy of Mandarin-accented speech by native English listeners over five sessions spanning 3 weeks. We found that the recognition of an accent feature (a syllable-final /d/, as in feed, sounding/t/-like) improved steadily over the three-week period. Unexpectedly, however, the improvement was seen with or without exposure to the accent. We discuss possible reasons for this result and implications for conducting future longitudinal studies with repeated exposure and testing.

Символический компонент семантики номинаций одежды (на материале говоров Среднего Приобья)

The article examines the symbolic component of the semantics of nominations for items of clothing in Russian traditional costume. The symbolic component of the meaning of the analyzed lexical units was identified through linguoculturological analysis. With its help, representatives are identified that testify to it as a cultural sign. The purpose of linguocultural analysis is to determine the totality of connotative components of the meaning of a word, including the symbolic one. To optimally structure a commentary, an answer to several questions is necessary: identifying the degree of objectivity of the procedure, clarifying the scope of its content, as well as determining the sequence of inclusion of components through which the symbolic semantics of a word is represented. The aim of the work is to determine the specifics of objectification of the cultural-symbolic component of the semantics of clothing nominations. The objectives are to identify lexical representative units that have a symbolic component of meaning in their semantics and to determine the set of symbolic components of meaning in clothing names. The research material was data from the dialects of the Middle Ob region recorded in the dictionaries of the Tomsk Dialectological School and the database of the Tomsk Dialect Corpus. The empirical basis of the work was formed using a continuous sampling method. The article implements the general scientific method of description. Methods of interpretation and linguocultural commentary are used. Techniques of contextual analysis, classification, systematization are the main ones when working with the studied layer of lexical units. The illustrative material included in the article contains facts of actualization of the symbolic component of clothing nominations in living dialect speech. Linguoculturological analysis, which allows stating that there is a symbolic connotative component in the semantics of the studied units, shows the idea of the symbolic in culture, which is firmly rooted in the archaic consciousness because it establishes a model of integral existence, knowledge about the orderliness of the world on the principles of correlation between phenomena and processes.