Automatic Speech Recognition Systems Research Articles

Automatic Speech Recognition System to Record Progress Notes in a Mobile EHR: A Pilot Study.

Creating notes in the EHR is one of the most problematic aspects for health professionals. The main challenges are the time spent on this task and the quality of the records. Automatic speech recognition technologies aim to facilitate clinical documentation for users, optimizing their workflow. In our hospital, we internally developed an automatic speech recognition system (ASR) to record progress notes in a mobile EHR. The objective of this article is to describe the pilot study carried out to evaluate the implementation of ASR to record progress notes in a mobile EHR application. As a result, the specialty that used ASR the most was Home Medicine. The lack of access to a computer at the time of care and the need to perform short and fast evolutions were the main reasons for users to use the system.

Development of an ASR System for Medical Conversations.

In this work we document the development of an ASR system for the transcription of conversations between patient and doctor and we will point out the critical aspects of the domain. The system was trained with an acoustic base of spontaneous speech that has a domain language model and a supervised phonetic dictionary. Its performance was compared with two systems: a) NeMo End-to-End Conformers in Spanish and b) Google API ASR (Automatic Speech Recognition) Cloud. The evaluation was carried out on a set of 208 teleconsultations recorded during the year 2020. The WER (Word Error Rate) was evaluated in ASR, and Recall and F1 for recognized medical entities. In conclusion, the developed system performed better, reaching 72.5% accuracy in the domain of teleconsultations and an F1 for entity recognition of 0.80.

Useful blunders: Can automated speech recognition errors improve downstream dementia classification?

Objectives:We aimed to investigate how errors from automatic speech recognition (ASR) systems affect dementia classification accuracy, specifically in the “Cookie Theft” picture description task. We aimed to assess whether imperfect ASR-generated transcripts could provide valuable information for distinguishing between language samples from cognitively healthy individuals and those with Alzheimer’s disease (AD). Methods:We conducted experiments using various ASR models, refining their transcripts with post-editing techniques. Both these imperfect ASR transcripts and manually transcribed ones were used as inputs for the downstream dementia classification. We conducted comprehensive error analysis to compare model performance and assess ASR-generated transcript effectiveness in dementia classification. Results:Imperfect ASR-generated transcripts surprisingly outperformed manual transcription for distinguishing between individuals with AD and those without in the “Cookie Theft” task. These ASR-based models surpassed the previous state-of-the-art approach, indicating that ASR errors may contain valuable cues related to dementia. The synergy between ASR and classification models improved overall accuracy in dementia classification. Conclusion:Imperfect ASR transcripts effectively capture linguistic anomalies linked to dementia, improving accuracy in classification tasks. This synergy between ASR and classification models underscores ASR’s potential as a valuable tool in assessing cognitive impairment and related clinical applications.

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.

Automated Measures of Syntactic Complexity in Natural Speech Production: Older and Younger Adults as a Case Study.

Multiple methods have been suggested for quantifying syntactic complexity in speech. We compared eight automated syntactic complexity metrics to determine which best captured verified syntactic differences between old and young adults. We used natural speech samples produced in a picture description task by younger (n = 76, ages 18-22 years) and older (n = 36, ages 53-89 years) healthy participants, manually transcribed and segmented into sentences. We manually verified that older participants produced fewer complex structures. We developed a metric of syntactic complexity using automatically extracted syntactic structures as features in a multidimensional metric. We compared our metric to seven other metrics: Yngve score, Frazier score, Frazier-Roark score, developmental level, syntactic frequency, mean dependency distance, and sentence length. We examined the success of each metric in identifying the age group using logistic regression models. We repeated the analysis with automatic transcription and segmentation using an automatic speech recognition (ASR) system. Our multidimensional metric was successful in predicting age group (area under the curve [AUC] = 0.87), and it performed better than the other metrics. High AUCs were also achieved by the Yngve score (0.84) and sentence length (0.84). However, in a fully automated pipeline with ASR, the performance of these two metrics dropped (to 0.73 and 0.46, respectively), while the performance of the multidimensional metric remained relatively high (0.81). Syntactic complexity in spontaneous speech can be quantified by directly assessing syntactic structures and considering them in a multivariable manner. It can be derived automatically, saving considerable time and effort compared to manually analyzing large-scale corpora, while maintaining high face validity and robustness. https://doi.org/10.23641/asha.24964179.

Аналитический обзор методов автоматического анализа экстралингвистических компонентов спонтанной речи

The accuracy of automatic spontaneous speech recognition systems is far from that of trained speech recognition systems. This is due to the fact that spontaneous speech is not as smooth and failure-free as spontaneous speech. Spontaneous speech varies from speaker to speaker: the quality of phonemes’ pronunciation, the presence of pauses, speech disruptions and extralinguistic items (laughing, coughing, sneezing, and chuckling when expressing emotions of irritation, etc.) interrupt the fluency of verbal speech. However, it is worth noting that extralinguistic items very often carry important paralinguistic information, so it is crucial for automatic spontaneous speech recognition systems not only to identify such phenomena and distinguish them from the verbal components of speech but also to classify them. This review presents an analysis of works on the topic of automatic detection and analysis of extralinguistic items in spontaneous speech. Both individual methods and approaches to the recognition of extralinguistic items in a speech stream, and works related to the multiclass classification of isolatedly recorded extralinguistic units are considered and described. The most popular methods of extralinguistic units’ analysis are neural networks, such as deep neural networks and networks based on transformer models. The basic concepts related to the term extralinguistic items are given, the original systematization of extralinguistic items in the Russian language is proposed, the corpus and databases of audio spoken speech both in Russian and in other languages are described, the data sets of extralinguistic items recorded isolatedly are also given. The accuracy of extralinguistic items recognition increases with the following conditions of work with the speech signal: pre-processing of audio signals of items has shown an increase in the accuracy of separately recorded extralinguistic items classification; consideration of context (analysis of several frames of speech signal) and use of filters for smoothing the time series after extraction of feature vectors showed an increase in accuracy in frame-by-frame analysis of the speech signal with spontaneous speech.

Developing children's ASR system under low-resource conditions using end-to-end architecture

The work presented in this paper aims at enhancing the performance of end-to-end (E2E) speech recognition task for children's speech under low resource conditions. For majority of the languages, there is hardly any speech data from child speakers. Furthermore, even the available children's speech corpora are limited in terms of the number of hours of data. On the other hand, large amounts of adults' speech data are freely available for research as well as commercial purposes. As a consequence, developing an effective E2E automatic speech recognition (ASR) system for children becomes a very challenging task. One may develop an ASR system using adults' speech and then use it to transcribe children's data, but this leads to very poor recognition rates due to the stark differences in the acoustic attributes of adults' and children's speech. In order to overcome these hurdles and to develop a robust children's ASR system employing E2E architecture, we have resorted to several out-of-domain and in-domain data augmentation techniques. For out-of-domain data augmentation, we have explicitly modified adults' speech to render it acoustically similar to that of children's speech before pooling into training. On the other hand, in the case of in-domain data augmentation, we have slightly modified the pitch and duration of children's speech in order to create more data capturing greater diversity. Data augmentation approaches helps in mitigating the ill-effects resulting from the scarcity of data from child domain to a certain extent. This, in turn, reduces the error rates by a large margin. In addition to data augmentation, we have also studied the efficacy of Gamma-tone frequency cepstral coefficients (GFCC) and frequency domain linear prediction (FDLP) technique along with the most commonly used Mel-frequency cepstral coefficients (MFCC) for front-end speech parameterization. Both MFCC as well as GFCC capture and model the spectral envelope of speech. On the other hand, application of linear prediction on the frequency domain representation of speech signal helps to effectively capture the temporal envelope during front-end feature extraction. Employing FDLP features that model the temporal envelope provides important cues for the perception and understanding of stop bursts and, at times, complete phonemes. This motivated us to perform a comparative experimental study of the effectiveness of the three aforementioned front-end acoustic features. In our experimental explorations, the use of proposed data augmentation in combination of FDLP features has shown a relative improvement in character error rate by 67.6% over the baseline system. The combination of data augmentation with MFCC or GFCC features is observed to result in lower recognition performances.

Using Voice Technologies to Support Disabled People

In recent years, significant strides have been made in speech and speaker recognition systems, owing to the rapid evolution of data processing capabilities. Utilizing a speech recognition system facilitates straightforward and efficient interaction, especially for individuals with disabilities. This article introduces an automatic speech recognition (ASR) system designed for seamless adaptation across diverse platforms. The model is meticulously described, emphasizing clarity and detail to ensure reproducibility for researchers advancing in this field. The model’s architecture encompasses four stages: data acquisition, preprocessing, feature extraction, and pattern recognition. Comprehensive insights into the system’s functionality are provided in the Experiments and Results section. In this study, an ASR system is introduced as a valuable addition to the advancement of educational platforms, enhancing accessibility for individuals with visual disabilities. While the achieved recognition accuracy levels are promising, they may not match those of certain commercial systems. Nevertheless, the proposed model offers a cost-effective solution with low computational requirements. It seamlessly integrates with various platforms, facilitates straightforward modifications for developers, and can be tailored to the specific needs of individual users. Additionally, the system allows for the effortless inclusion of new words in its database through a single recording process.

An investigation and analysis on automatic speech recognition systems

<p>A crucial part of a Speech Recognition System (SRS) is working on its most fundamental modules with the latest technology. While the fundamentals provide basic insights into the system, the recent technologies used on it would provide more ways of exploring and exploiting the fundamentals to upgrade the system itself. These upgrades end up in finding more specific ways to enhance the scope of SRS. Algorithms like the Hidden Markov Model (HMM), Artificial Neural Network (ANN), the hybrid versions of HMM and ANN, Recurrent Neural Networks (RNN), and many similar are used in accomplishing high performance in SRS systems. Considering the domain of application of SRS, the algorithm selection criteria play a critical role in enhancing the performance of SRS. The algorithm chosen for SRS should finally work in hand with the language model conformed to the natural language constraints. Each language model follows a variety of methods according to the application domain. Hybrid constraints are considered in the case of geography-specific dialects.</p>

Linguistic disparities in cross-language automatic speech recognition transfer from Arabic to Tashlhiyt

Tashlhiyt is a low-resource language with respect to acoustic databases, language corpora, and speech technology tools, such as Automatic Speech Recognition (ASR) systems. This study investigates whether a method of cross-language re-use of ASR is viable for Tashlhiyt from an existing commercially-available system built for Arabic. The source and target language in this case have similar phonological inventories, but Tashlhiyt permits typologically rare phonological patterns, including vowelless words, while Arabic does not. We find systematic disparities in ASR transfer performance (measured as word error rate (WER) and Levenshtein distance) for Tashlhiyt across word forms and speaking style variation. Overall, performance was worse for casual speaking modes across the board. In clear speech, performance was lower for vowelless than for voweled words. These results highlight systematic speaking mode- and phonotactic-disparities in cross-language ASR transfer. They also indicate that linguistically-informed approaches to ASR re-use can provide more effective ways to adapt existing speech technology tools for low resource languages, especially when they contain typologically rare structures. The study also speaks to issues of linguistic disparities in ASR and speech technology more broadly. It can also contribute to understanding the extent to which machines are similar to, or different from, humans in mapping the acoustic signal to discrete linguistic representations.

ArzEn-LLM: Code-Switched Egyptian Arabic-English Translation and Speech Recognition Using LLMs

Motivated by the widespread increase in the phenomenon of code-switching between Egyptian Arabic and English in recent times, this paper explores the intricacies of machine translation (MT) and automatic speech recognition (ASR) systems, focusing on translating code-switched Egyptian Arabic-English to either English or Egyptian Arabic. Our goal is to present the methodologies employed in developing these systems, utilizing large language models such as LLama and Gemma. In the field of ASR, we explore the utilization of the Whisper model for code-switched Egyptian Arabic recognition, detailing our experimental procedures including data preprocessing and training techniques. Through the implementation of a consecutive speech-to-text translation system that integrates ASR with MT, we aim to overcome challenges posed by limited resources and the unique characteristics of the Egyptian Arabic dialect. Evaluation against established metrics showcases promising results, with our methodologies yielding a significant improvement of 56% in English translation over the state-of-the-art and 9.3% in Arabic translation. Since code-switching is deeply inherent in spoken languages, it is crucial that ASR systems can effectively handle this phenomenon. This capability is crucial for enabling seamless interaction in various domains, including business negotiations, cultural exchanges, and academic discourse. Our models and code are available as open-source resources.1,2.

Real-Time Automatic Continuous Speech Recognition System for Kannada Language/Dialects

Real-Time Automatic Continuous Speech Recognition System for Kannada Language/Dialects

Training speech recognition models at the National Library of Sweden

The abundance of openly available audio data in English enables pretraining of automatic speech recognition (ASR) systems on hundreds of thousands to millions of hours of recorded speech. As a result speech recognition systems are approaching human level robustness in English. Other languages’ performance in multilingual speech recognition systems tend to stand in proportion to the amount of data included from the language – or the language family – in question. For low to mid resource languages with fewer speakers, the amount of openly available data may be limited, and as a consequence these languages tend to be underrepresented in large scale efforts to train multilingual speech recognition systems. The National Library of Sweden hosts large collections of audio recordings. These resources allow us to potentially bridge some of the existing speech recognition performance gaps between Swedish and higher resource languages. We believe Swedish speech recognition can be further improved upon by scaling up the amount of training data. We have constructed an inclusive and transparent speech corpus with emphasis on all variations of spoken Swedish that we will use to train speech-to-text models for Swedish.

Adapting Pre-Trained Self-Supervised Learning Model for Speech Recognition with Light-Weight Adapters

Self-supervised learning (SSL) is an effective way of learning rich and transferable speech representations from unlabeled data to benefit downstream tasks. However, effectively incorporating a pre-trained SSL model into an automatic speech recognition (ASR) system remains challenging. In this paper, we propose a network architecture with light-weight adapters to adapt a pre-trained SSL model for an end-to-end (E2E) ASR. An adapter is introduced in each SSL network layer and trained on the downstream ASR task, while the parameters of the pre-trained SSL network layers remain unchanged. By carrying over all pre-trained parameters, we avoid the catastrophic forgetting problem. At the same time, we allow the network to quickly adapt to ASR task with light-weight adapters. The experiments using LibriSpeech and Wall Street Journal (WSJ) datasets show that (1) the proposed adapter-based fine-tuning consistently outperforms full-fledged training in low-resource scenarios, with up to 17.5%/12.2% relative word error rate (WER) reduction on the 10 min LibriSpeech split; (2) the adapter-based adaptation also shows competitive performance in high-resource scenarios, which further validates the effectiveness of the adapters.

Legendas em árabe geradas por inteligência artificial: insights do sistema de reconhecimento automático de fala do árabe jordaniano da Veed.io

Abstract This paper examines the errors that the automatic speech recognition (ASR) system of Veed.io produces when transcribing utterances spoken in Jordanian Arabic into subtitles. It attempts to propose a new classification for the subtitles that are built based on artificial intelligence technology. Through a combination of qualitative and quantitative analyses, the study examines the types of errors and their impact on comprehension. The errors observed in the generated subtitles based on linguistic and phonetic analysis are categorised into three main types: deletions, substitutions, and insertions. Furthermore, the quantitative analysis measures the word error rate (WER) and shows that the WER percentage is 38.857% revealing that deletions are the most common type of error, followed by substitutions and insertions. The study recommends conducting further research on ASR systems for Arabic language dialects and advises subtitlers to be aware of the limitations of these systems when using them, ensuring that they edit and supervise them appropriately.

基于喉部振动的语音自动识别系统的设计

基于喉部振动的语音自动识别系统的设计

Enhancing Armenian Automatic Speech Recognition Performance: A Comprehensive Strategy for Speed, Accuracy, and Linguistic Refinement

This research introduces a comprehensive strategy to enhance the performance of an existing automatic speech recognition (ASR) model, which has been previously documented in published articles. The study sets out to achieve several objectives. Firstly, it concentrates on updating the ASR model by retraining it with new datasets. This involves integrating samples from the latest Common Voice corpus release and data collected independently via the armspeech.com web application. Another key focus lies in optimizing the ASR model for near-real-time processing, intending to improve its speed and efficiency. The proposed adjustments to the model’s architecture aim to balance accuracy and processing speed, which is essential for applications requiring prompt speech recognition. Furthermore, the research explores the integration of Transformer models into the post-processing pipeline to introduce punctuation and capitalization into the ASR output. This step not only enhances the linguistic quality of transcriptions but also improves their readability and usability. In tandem with these advancements, the research presents a systematic approach to gathering, annotating, and storing datasets specifically tailored for punctuation and capitalization tasks. The methodology outlines the acquisition and organization of a dataset conducive to training Transformer models for these linguistic tasks. This comprehensive approach, which encompasses dataset enrichment, architectural modifications, and post-processing enhancements, aims to elevate the ASR model’s accuracy, speed, and linguistic refinement, with a particular focus on addressing the intricacies of the Armenian language. The research contributes valuable insights into the optimization of ASR systems, tackling both language-specific challenges and broader issues related to linguistic post-processing.

Toward Effective Aircraft Call Sign Detection Using Fuzzy String-Matching between ASR and ADS-B Data

Recently, artificial intelligence and data science have witnessed dramatic progress and rapid growth, especially Automatic Speech Recognition (ASR) technology based on Hidden Markov Models (HMMs) and Deep Neural Networks (DNNs). Consequently, new end-to-end Recurrent Neural Network (RNN) toolkits were developed with higher speed and accuracy that can often achieve a Word Error Rate (WER) below 10%. These toolkits can nowadays be deployed, for instance, within aircraft cockpits and Air Traffic Control (ATC) systems in order to identify aircraft and display recognized voice messages related to flight data, especially for airports not equipped with radar. Hence, the performance of air traffic controllers and pilots can ultimately be improved by reducing workload and stress and enforcing safety standards. Our experiment conducted at Tangier’s International Airport ATC aimed to build an ASR model that is able to recognize aircraft call signs in a fast and accurate way. The acoustic and linguistic models were trained on the Ibn Battouta Speech Corpus (IBSC), resulting in an unprecedented speech dataset with approved transcription that includes real weather aerodrome observation data and flight information with a call sign captured by an ADS-B receiver. All of these data were synchronized with voice recordings in a structured format. We calculated the WER to evaluate the model’s accuracy and compared different methods of dataset training for model building and adaptation. Despite the high interference in the VHF radio communication channel and fast-speaking conditions that increased the WER level to 20%, our standalone and low-cost ASR system with a trained RNN model, supported by the Deep Speech toolkit, was able to achieve call sign detection rate scores up to 96% in air traffic controller messages and 90% in pilot messages while displaying related flight information from ADS-B data using the Fuzzy string-matching algorithm.

Accents in Speech Recognition through the Lens of a World Englishes Evaluation Set

Automatic Speech Recognition (ASR) systems generalize poorly on accented speech, creating bias issues for users and providers. The phonetic and linguistic variability of accents present challenges for ASR systems in both data collection and modeling strategies. We present two promising approaches to accented speech recognition— custom vocabulary and multilingual modeling— and highlight key challenges in the space. Among these, lack of a standard benchmark makes research and comparison difficult. We address this with a novel corpus of accented speech: Earnings-22, A 125 file, 119 hour corpus of English-language earnings calls gathered from global companies. We compare commercial models showing variation in performance when taking country of origin into consideration and demonstrate targeted improvements using the methods we introduce.

AfriSpeech-200: Pan-African Accented Speech Dataset for Clinical and General Domain ASR

Abstract Africa has a very poor doctor-to-patient ratio. At very busy clinics, doctors could see 30+ patients per day—a heavy patient burden compared with developed countries—but productivity tools such as clinical automatic speech recognition (ASR) are lacking for these overworked clinicians. However, clinical ASR is mature, even ubiquitous, in developed nations, and clinician-reported performance of commercial clinical ASR systems is generally satisfactory. Furthermore, the recent performance of general domain ASR is approaching human accuracy. However, several gaps exist. Several publications have highlighted racial bias with speech-to-text algorithms and performance on minority accents lags significantly. To our knowledge, there is no publicly available research or benchmark on accented African clinical ASR, and speech data is non-existent for the majority of African accents. We release AfriSpeech, 200hrs of Pan-African English speech, 67,577 clips from 2,463 unique speakers across 120 indigenous accents from 13 countries for clinical and general domain ASR, a benchmark test set, with publicly available pre-trained models with SOTA performance on the AfriSpeech benchmark.