In this paper, an assessment of the combined effects of the noise, early and late reflections on speech (monosyllables of consonance-vowel-consonance type) intelligibility for binaural listening mode was made. Test signals were synthesized in two ways. In the first case, the noisy monosyllables (signal-to-noise ratio was varied from −15 dB to +5 dB) was filtered by two-channel room impulse responses for distances varied from 2 m to 10 m from the sound source in the middle size auditorium. In the second case, the room impulse responses were cut after 50 ms and used for filtering the same noised monosyllables. Filtered monosyllables were listened through headphones and entered into the computer using the keyboard. The results of testing showed that monosyllables intelligibility is the highest for 2 m distance and is the lowest in the middle of the room. A slight increase in speech intelligibility at the back wall can be explained by the combined action of direct sound and early reflections. The decrease in speech intelligibility in the middle of the room can be further explained by a decrease in the quality of the speech signal.

The Effects of Noise, Early and Late Reflections on Speech Intelligibility

The proposed technique of objective (without patient’s participation) audiometry allows measuring hearing loss. It is based on a developed model of a middle ear, obtained with its model the formula for a individual middle ear norm parameter. On the basis of the formula the resonant coefficients of transformation of pressure through ossicles and amplification of pressure in auditory meatus are designed.

Objective Audiometry Based on the Formula of the Middle Ear Parameter: A New Technique for Researches and Differential Diagnosis of Hearing

Deep learning-based speech enhancement algorithms have shown their powerful ability in removing both stationary and non-stationary noise components from noisy speech observations. But they often introduce artificial residual noise, especially when the training target does not contain the phase information, e.g., ideal ratio mask, or the clean speech magnitude and its variations. It is well-known that once the power of the residual noise components exceeds the noise masking threshold of the human auditory system, the perceptual speech quality may degrade. One intuitive way is to further suppress the residual noise components by a postprocessing scheme. However, the highly non-stationary nature of this kind of residual noise makes the noise power spectral density (PSD) estimation a challenging problem. To solve this problem, the paper proposes three strategies to estimate the noise PSD frame by frame, and then the residual noise can be removed effectively by applying a gain function based on the decision-directed approach. The objective measurement results show that the proposed postfiltering strategies outperform the conventional postfilter in terms of segmental signal-to-noise ratio (SNR) as well as speech quality improvement. Moreover, the AB subjective listening test shows that the preference percentages of the proposed strategies are over 60%.

/pdf/low-complexity-artificial-noise-suppression-methods-for-deep-4qf8cq8ir3.pdf

Low-complexity artificial noise suppression methods for deep learning-based speech enhancement algorithms

The intelligibility assessment of dysarthric speech is essential for planning therapy. Time-frequency representations have been used for automatic intelligibility assessment of dysarthria. These representations have been derived from the utterance as a whole. As voiced and unvoiced components have different characteristics; in this work, we use different time-frequency representations for voiced and unvoiced segments and use them for intelligibility assessment with CNN classifier. Finally, we combine the scores obtained by the two systems to assign an intelligibility level. The combined system’s performance is found to be superior to the systems that used both voiced and unvoiced components separately or together as one utterance. The utterances of the TORGO database are used in the intelligibility assessment. Automatic assessment of speech intelligibility reduces speech-language pathologists’ time and effort in assisting diagnosis and treatment design.

Region Based Prediction and Score Combination for Automatic Intelligibility Assessment of Dysarthric Speech

In this paper, the technology of correction of the frequency response of the measuring path of the hardware-software complex "Artificial head", intended for acoustic examination of the rooms, has been developed. This correction is necessary because the amplitude-frequency response of the loudspeaker and the microphone are not perfectly uniform in the frequency band of analysis. Therefore, instead of the room impulse response, the convolution of the room impulse response with the impulse responses of the loudspeaker (test signal source) and the microphone will actually be evaluated. It is shown that such correction can be made by controlled dividing the frequency response of the loudspeaker-room-microphone system by the previously obtained estimate of the amplitude-frequency response of the loudspeaker-microphone subsystem. The problem with such calculations is the division operation, because the amplitude-frequency response of the loudspeaker-microphone subsystem may contain small numerical values, which will overflow the discharge grid of the computing system and crash the computer. However, it is clear that if proper control over the amplitude-frequency characteristics of the loudspeaker-microphone subsystem is ensured, such a division can be practically implemented. Since the amplitude-frequency response of the subsystem "loudspeaker-microphone" takes the smallest values at the edges of the frequency range, and the variance of the evaluation of the mutual spectrum of the system "loudspeaker-room-microphone" is the largest on the right edge of the frequency range, it is advisable to apply the method of regularization to achieve necessary calculation accuracy. In this case, the role of the regularizing factor can be played by a spectral weighting window whose values are close to one at low and medium frequencies, and with the approaching to the right edge of the frequency range, the values of this window are close to zero. The role of the regularization parameter is thus the width of such a weighting window. The nature and value of such correction effect on the accuracy of the room impulse response estimate are analyzed in this paper. It is shown that the Hann (Hanning) window can be used as a regularizing factor, while the width value of this window was experimentally found, which provides satisfactory properties for evaluating the room impulse response. It has been also shown that the width of the Hann window, which is close to 80% of the total frequency range analyzed, is satisfactory for practical applications.

/pdf/aparatno-programnii-kompleks-shtuchna-golova-chastina-1-318lbsl5mx.pdf

Апаратно-програмний комплекс «Штучна голова». Частина 1 Коригування частотної характеристики тракту

Comparison of the behavior of subjective assessments of quality and intelligibility of speech distorted by white, pink and brown noise is performed. The results obtained indicate the similarity of this behavior in the range of signal-to-noise (SNR) values 0 0 dB, whereas the tendency for such agreement is observed only for SNR>10 dB for frequency-weighted segmental SNR (fwSNR) measure.

Objective and Subjective Assessment of the Quality and Intelligibility of Noised Speech

In this paper, subjective and objective estimators of the quality of speech and music signals subjected to phase distortion are compared, and mapping between objective and subjective quality estimates is realized. It was found that the phase distortion of speech signals is perceived stronger than ones for musical signals. Two types of phase distortion are considered: 1) low-frequency signal components lag behind high-frequency components in the 30–90 ms; 2) high-frequency signal components lag behind low frequency components in the same time. It is shown that the human auditory system is more sensitive to the latter type of phase distortion referenced previously. As far as objective quality measures, it was found they are useless for distinguishing these types of phase distortion. Maximum differences between group delay times for low (125 Hz) and high (8 kHz) frequencies, for which distortion is acceptable, were established. These threshold values are 40 ms for speech and 80 ms for music.

On peculiarities of evaluating the quality of speech and music signals subjected to phase distortion

In this paper, the results of experimental studies of the reasons for the low reliability of the log-spectral distortion (LSD) measure for estimating the quality of speech signals limited in the frequency band are presented. It is shown that the LSD measure has an increased sensitivity not only to the speech formants, but also to the shape of frequency response of low-frequency filters limiting the frequency band. It is also shown that normalization by variance of band-limited signals leads to LSD measure anomalous behavior. Simple modification of the LSD estimate is proposed, which makes it possible to increase its statistical stability.

On reliability of log-spectral distortion measure in speech quality estimation

When transmitting music signals via infocommunication channels, there is a risk of distortion of these signals due to clipping. The consequence of these distortions is a decrease in subjective assessments of the quality of musical signals, so clipping detection is an important issue. In this paper, the kurtosis and the inverse value of kurtosis are proposed as measures of an objective assessment of the musical signals clipping degree. Matching maps of objective and subjective clipping degree estimates are monotonic functions that do not contain sharp kinks, which allow not only evaluating clipping degree, but also estimating music signals quality. An important advantage of the proposed measures of the clipping degree is the ability to simplify and speed up significantly the calculations due to the absence of the need of preliminarily estimation the probability density function of the analyzed musical signal.

Using Kurtosis for Objective Assessment of the Musical Signals Clipping Degree

Clipping of speech leads to the appearance of higher orders harmonics and, as a result, to reducing of accuracy of automatic speech recognition systems used as kind of artificial intellect of aircraft control systems and flight control systems for unmanned aerial vehicles. In this paper, the subjective and objective estimates of clipped speech quality are presented. It was shown that subjective speech quality Degradation Mean Opinion Score scale values are about 4.5, 3.5 and 2.5 for degrees of clipping 5 dB, 10 dB and 15 dB, respectively. Establishing of this rule allows find the boundary permissible degree of clipping based on certain requirements to the speech quality. Dependencies of objective speech quality measures such as segmental signal-to-noise ratio, frequency weighted segmental signal-to-noise ratio, log-spectral distortion, bark-spectral distortion and perceptual evaluation of speech quality on the clipping degree are obtained. Calculations of correlation coefficients and matching maps which establish relationship between objective and subjective speech quality measures have been made. Obtained results allow concluding that objective speech quality measures can be applied to evaluate both the clipped speech quality and the degree of speech signals clipping.

Clipped Speech Signals Quality Estimation

Handle everyday research tasks with reliable, citation-backed results

Your personal Research Agent to handle research tasks with citation-backed results

Popular Tasks used by Researchers

How can I help with your research?

Meet SciSpace

Get more enhanced response by uploading the PDFs you want me to reference.

No relevant PDFs in your library

SciSpace is the AI research assistant for academics. Run systematic literature reviews on 280M+ papers, and write papers with cited sources. Trusted by 1M+ students, PhDs & researchers.

SciSpace | AI for Research

Analyze PDFs

Code & Manuscripts

Funding & Grants

Literature & Patents

Medical & Clinical Data

Systematic Review

Visualize & Present

Web & Data

Build a Google Scholar-like website for your research.

Build a website

Create charts and images for your research

Create a Chart

Write a paper for submission to a journal

Draft a manuscript

Patent Search

Design eye-catching scientific posters in minutes.

Scientific Poster Generation

Systematic Literature Review

One task is running at the moment. Your messages will be shown right after.

Drag and drop or click here to browse

Loved by <highlight>1 million+</highlight> researchers

Extract a list of specific topics and their sources from unstructured text

Topics

Compare and analyze relevant papers that matches with your search

Papers

Get insights from PDFs and bookmarked papers from your library

My library

Recent searches

Try searching for:

Catch AI-generated content in scholarly and non-scholarly content

{ai} Detector

Ai Writer

Get PDF Summaries, highlighted text explanations 

Chat with PDF

Effortlessly create in-text citations and bibliographies in APA and 2,500 other formats

Citation generator

Get explanations, summaries, and answers on academic papers

Ease up your research workflow with {scispace}'s cohort of exciting AI tools

Elevate your academic writing skills and convey your ideas the way you want

Paraphraser

Explore our range of reading and writing tools

Your file is being prepared and should be ready in a few minutes. If it's a large file, it might take a bit longer. You can close this window, and we'll email you the file when it's done.

You have reached a maximum limit of <strong>{limit}</strong> columns in the table. Remove at least <strong>1</strong> column to add or create another one.

Igor Kotvytskyi

Author Tools

Chat about Author