speech recognition

The output Signal-to-Noise Ratio (SNR) is one of the essential factors in hearing aid benefits. There is limited evidence regarding SNR improvement by the Channel-Free (CFHA) and Multi-Channel Hearing Aid (MCHA) and the speech understanding in noise through them. This study aimed to investigate the extent to which output SNR was modified by CFHA and MCHA processing and the variation in aided speech recognition abilities with a change in output SNR.

Thirty-six participants aged 50–65 years were included. A chosen CFHA and MCHA were used to obtain the output SNR and sentence recognition in noise in four different processing algorithms (linear, linear+noise reduction, WDRC, WDRC+noise reduction). Hagerman’s phase inversion technique was used to measure the attenuation of noise and, in turn, to obtain the output SNR of the hearing aid.

In all hearing aid processing algorithms among those with normal hearing and people with hearing loss, the output of CFHA revealed higher attenuation values than that of MCHA. There was a significant effect of the hearing aids and processing algorithms in both normal and individuals with hearing impairment on the mean SNR. Further, multiple linear regression analysis results showed that whether the hearing is channel-free or multichannel significantly predicted speech recognition scores, while output SNR and processing algorithms did not.

The signal processing algorithms in CFHA had greater noise attenuation values, better output SNR, and speech recognition scores, showing an advantage over the modern MCHA among individuals with hearing impairment.

Using personal protective equipment, especially face masks, in the COVID-19 pandemic era may make verbal communication difficult. Furthermore, acoustic changes in mask-wearing conditions may affect speech audiometry results. This study investigates the effect of wearing a face mask on the word recognition score and the role of speech frequency content in this effect.

This study was planned and conducted in two phases. In the first phase, the validity and reliability of two speech material lists, high-pitch and low-pitch, were determined. In the second phase, the word recognition score was measured for “mask-wearing” and “covering mouth without a mask.”

The statistical analysis showed that the content validity ratio was 0.92, and the content validity index was 0.8. Therefore, both speech lists were valid. For these lists, the mixed analysis of variance analysis showed that the scores for “mask-wearing” were significantly lower than “covering mouth without a mask,” and there was more reduction in scores for the high-pitched list (F=8.7, df=1, P<0.005).

In terms of the impact of a face mask on speech, explaining how speech audiometry is performed, especially in monitoring treatments, may help limit the probability of misinterpretation of speech test findings. Furthermore, understanding the impact of face masks on word recognition scores in adopting sufficient auditory rehabilitation procedures is necessary.

Identifying sound localization defects in children and using appropriate rehabilitation methods lead to improve their problem to use binaural processing skills to successfully suppress background noise and selectively attend to a particular auditory source. This pilot study aims to design a game for auditory training of children with Spatial Processing Disorder (SPD) in Iran and compare their performance in engagement time and speech recognition in noise with those of normal peers.

The game application was designed based on the spatial separation of target and competing stimuli by applying head-related transfer functions. The child’s task was to identify the picture of the target word from among the displayed options. Participants were 24 children with SPD and 27 normal peers aged 7–9 years, who performed the speech recognition tasks using the designed game.

The game was fascinating and easy to use for most participants. There was no significant difference in mean engagement time between the two groups. The mean engagement time of participants was significantly different between the two groups in terms of age. The children with SPD obtained lower scores in speech recognition in noise tasks. There was no significant difference in speech recognition scores between girls and boys.

The designed game application has the potential for speech recognition innoise training of children with SPD in Iran. It is a fascinating and user-friendly tool for simulating the real-life auditory situations for these children.

Background :

Speech recognition is a significant component of speech perception assessment as a main clinical target in children with hearing disorders. The current article presents a systematic review designed to provide clinical and research guidelines for speech recognition assessment in Persian-speaking children with and without hearing disorders.

A systematic search was conducted to cumulate the research evidence for the assessment of speech recognition in Persian-speaking children with and without hearing disorders. Peer-reviewed journal articles dedicated to this issue and published between 1982 and December 2021 were discovered through a search of the electronic databases of PubMed, Scopus, MEDLINE, Web of Science, Magiran, IranMedex, Scientific Information Database (SID), and Google Scholar. The papers were analyzed according to different variables, including test age, vocabulary competency, cognitive demands, response format, presentation mode, stimulus format, stimulus variability, stimulus mode, and test conditions.

The review identified four papers related to the development of five assessment tools for measuring speech recognition in Persian-speaking children, namely Persian Monosyllabic Lexical Neighborhood Tests (PMLNTs), Persian Disyllabic Lexical Neighborhood Tests (PDLNTs), Persian version of the words-in-noise (WIN), Tavana (test for evaluating auditory skills), and closed-set speech recognition test for Persian-speaking children. The properties and limitations of each test have been considered in the current review article.

Based on this systematic review, only four speech recognition assessment tools exist for measuring specifically spoken word recognition in Persian-speaking children. The results of this review article can be used as a clinical and research guideline for assessing speech recognition in Persian-speaking children with hearing disorders.

The use of cochlear implants, due to technological limitations, causes problems in speech comprehension in the presence of noise. This study aimed to evaluate the speech-in-noise (SIN) comprehension with emphasis on high-frequency components between users of different bimodal adult.

This study was conducted on 33 adult participants with a mean age of 36 years using bimodal (cochlear implant in one ear and hearing aid in another ear: CI/HA) style of different companies. Quick SIN with emphasis on high-frequency components was performed on the participants using an audiometer, an amplifier, and one speaker.

Comparing the average percentage of correct answers from the word recognition test in the presence of noise in bimodal users showed that the Cochlear brand provides a better signal-to-noise (SNR) compare to other brands. Our result shows that bimodal users of Advance bionic and Med-El groups have better performance in speech recognition than other brands.

Bimodal users of Advance bionic and Med-El have better SNR loss than other brands. Besides, further studies on different ages can be helpful to make the right decision in this regard.

In this paper, a novel approach is proposed for the recognition of Persian phonemes in the Persian consonant‑vowel combination (PCVC) speech dataset. Nowadays, deep neural networks (NNs) play a crucial role in classification tasks. However, the best results in speech recognition are not yet as perfect as human recognition rate. Deep learning techniques show outstanding performance over many other classification tasks, such as image classification and document classification. Furthermore, the performance is sometimes better than a human. The reason why automatic speech recognition systems are not as qualified as the human speech recognition system, mostly depends on features of data which are fed to deep NNs.

In this research, first, the sound samples are cut for the exact extraction of phoneme sounds in 50 ms samples. Then, phonemes are divided into 30 groups, containing 23 consonants, 6 vowels, and a silence phoneme.

The short‑time Fourier transform is conducted on them, and the results are given to PPNet (a new deep convolutional NN architecture) classifier and a total average of 75.87% accuracy is reached which is the best result ever compared to other algorithms on separated Persian phonemes (like in PCVC speech dataset).

This method not only can be used for recognizing mono‑phonemes but it can also be adopted as an input to the selection of the best words in speech transcription

Speech recognition (SR) technology has been existing for more than two decades. But, it has been rarely used in health care institutions and not applied uniformly in all the clinical domains. The aim of this study was to investigate the accuracy of speech recognition system in four different situations in the real environment of health services. We also report physicians' experience of using speech recognition technology.

To do this study, NEVISA SR software professional v.3 was installed on the computers of expert physicians. The pre-designated medical report was tested by the physicians in four different modes including slow expression in a silent environment, slow expression in crowded environments, rapid expression in a silent environment and rapid expression in a busy environment. After using the speech recognition software by 15 physicians in hospitals, a designed questionnaire was distributed among them.

The results showed that the highest average accuracy of speech recognition software was in the silent environment by slow expression and the minimum average accuracy was in the busy environment by rapid expression. Of all the participants in the study, 53.3% of the physicians believed that the use of speech recognition system promoted the workflow.

We found that software accuracy was generally higher than the expectation and its use required to upgrade the system and its operation. In order to achieve the highest level of recognition rate and error reduction by speech recognition, influential factors such as environmental noise, type of software or hardware, training and experience of participants can be also considered.

Bilingualism is an important phenomenon with different effects on each aspect of language processing. Auditory temporal processing is a major component of the auditory processing ability. Since bilingual and monolingual individual’s brain process are different, and no studies have yet been conducted on the effect of temporal processing on speech recognition performance of Azeri-Persian bilinguals, the purpose of the present study is the comparison of auditory temporal processing between Persian monolinguals and early Azeri-Persian bilinguals. In this cross-sectional study, the Persian version of time compressed speech test was performed with monosyllabic words in three compression rates of 0, 40 and 60 percent in the most comfortable level in each ear, separately. The subjects were 36 Persian monolinguals and 36 Azeri-Persian bilinguals aged from 18 to 30 and were analyzed based on languages’ compression rate and ear factors. The speech recognition scores (SRS) in both groups were decreased by increasing compression rate and a significant difference was shown between mean SRS of the two studied groups (p<0.0001). There was no significant difference between three compression rates by the ear (p>0.05). In compression rate of 40 percent, bilingual group had lower speech recognition ability in comparison with monolingual group. In addition, there was more significant difference when compression rate increases from 40 to 60 percent.

Fast and holistic access to the patients’ clinical record is a major requirement of modern medical decision support systems (DSS). While electronic health records (EHRs) have replaced the traditional paper-based records in most healthcare organization, the data entry into these systems remains largely manual. Speech recognition technology promises substitution of the more convenient speech-based data entry with currently laborious manual method, in the near future. Developing effective speech recognition systems (SRS) require availability of standardized vocabulary databases. This study was aimed at developing a medical speech recognition database for reconstructive hand surgery based on the Systematized Nomenclature of Medicine-Clinical Terms (SNOMED-CT). All codes related to hand problems were extracted from ICD-10. A sample of 2051 diagnosis codes was randomly selected from the patients’ records. The operation report paper sheets were transformed to electronic records. For each term, the SNOMED-CT was browsed to find the preferred synonym, using CliniClue® Xplore Software. For some words with several number of synonyms, the preference of reconstructive surgery specialists were asked using a researcher-made questionnaire. Ultimately, the preferred words was substituted throughout each document and used for developing a database of standard nomenclature. The developed database was used in speech-based recording clinical data in reconstructive hand surgery operating room and accuracy of 81% in correct recording of clinical data was observed. Therefore, development of standard medical nomenclature databases can facilitate accurate electronic recording of medical data and reduce the associated labor and cost posed by current manual method.