Perceptual Evaluation Of Quality Research Articles

Blind Quality Assessment of Dense 3D Point Clouds with Structure Guided Resampling

Objective quality assessment of 3D point clouds is essential for the development of immersive multimedia systems in real-world applications. Despite the success of perceptual quality evaluation for 2D images and videos, blind/no-reference metrics are still scarce for 3D point clouds with large-scale irregularly distributed 3D points. Therefore, in this paper, we propose an objective point cloud quality index with Structure Guided Resampling (SGR) to automatically evaluate the perceptually visual quality of dense 3D point clouds. The proposed SGR is a general-purpose blind quality assessment method without the assistance of any reference information. Specifically, considering that the human visual system (HVS) is highly sensitive to structure information, we first exploit the unique normal vectors of point clouds to execute regional pre-processing which consists of keypoint resampling and local region construction. Then, we extract three groups of quality-related features, including: 1) geometry density features; 2) color naturalness features; 3) angular consistency features. Both the cognitive peculiarities of the human brain and naturalness regularity are involved in the designed quality-aware features that can capture the most vital aspects of distorted 3D point clouds. Extensive experiments on several publicly available subjective point cloud quality databases validate that our proposed SGR can compete with state-of-the-art full-reference, reduced-reference, and no-reference quality assessment algorithms.

Gap-Closing Matters: Perceptual Quality Evaluation and Optimization of Low-Light Image Enhancement

Gap-Closing Matters: Perceptual Quality Evaluation and Optimization of Low-Light Image Enhancement

High-capacity steganography through audio fusion and fission

<span>Information security is required for two reasons, either to conceal the information completely or to prevent the misuse of the information by adding watermarks or metadata. Audio steganography uses audio signals to hide secret information. In the proposed audio steganography technique, cover audio files and secret audio files are transformed from time domain to wavelet domain using discrete wavelet transform, the secret audio file is transformed in two levels, leading to secure and high-capacity data hiding. 1% of the 2-level compressed secret is fused to 99% of the 1-level compressed cover. “Peak signal to noise ratio and mean squared error, Pearson’s correlation coefficient, spearman’s correlation coefficient, perceptual evaluation of speech quality and short-time objective intelligibility” are considered to assess the similarity of cover audio and stego audio and similarity of secret audio embedded, and secret audio retrieved. Results show that the stego audio signal is perceptually indistinguishable from the cover audio signal. The approach also passed the robustness test.</span>

Research on the Perceived Quality of Virtual Reality Headsets in Human-Computer Interaction.

The progress of commercial VR headsets largely depends on the progress of sensor technology, the iteration of which often means longer research and development cycles, and also higher costs. With the continuous maturity and increasing competition of VR headsets, designers need to create a balance among user needs, technologies, and costs to achieve commercial competition advantages. To make accurate judgments, consumer feedback and opinions are particularly important. Due to the increasing maturity in the technology of commercial VR headsets in recent years, the cost has been continuously decreasing, and potential consumers have gradually increased. With the increase in consumer demand for virtual reality headsets, it is particularly important to establish a perceptual quality evaluation system. The relationship between consumer perception and product quality determined by evaluations of experience is improving. Using the research method implemented in this work, through semi-structured interviews and big data analysis of VR headset consumption, the perceptual quality elements of VR headsets are proposed, and the order of importance of perceptual quality attributes is determined by questionnaire surveys, quantitative analysis, and verification. In this study, the perceptual quality elements, including technical perceptual quality (TPQ) and value perceptual quality (VPQ), of 14 types of VR headsets were obtained, and the importance ranking of the VR headsets' perceptual quality attributes was constructed. In theory, this study enriches the research on VR headsets. In practice, this study provides better guidance and suggestions for designing and producing VR headsets so that producers can better understand which sensor technology has met the needs of consumers, and which sensor technology still has room for improvement.

Speech Enhancement Based on Enhanced Empirical Wavelet Transform and Teager Energy Operator

This paper presents a new speech-enhancement approach based on an enhanced empirical wavelet transform, considering the time and scale adaptation of thresholds for individual component signals obtained from the used transform. The time adaptation is performed using the Teager energy operator on the individual component signals, and the scale adaptation of thresholds is performed by the modified level-dependent threshold principle for the individual component signals. The proposed approach does not require an explicit estimation of the noise level or a priori knowledge of the signal-to-noise ratio as is usually needed in most common speech-enhancement methods. The effectiveness of the proposed method has been assessed based on over 1000 speech recordings from the public Librispeech database. The research included various types of noise (among others white, violet, brown, blue, and pink) and various types of disturbance (among others traffic sounds, hair dryer, and fan), which were added to the selected test signals. The score of perceptual evaluation of speech quality, allowing for the assessment of the quality of enhanced speech, and signal-to-noise ratio, allowing for the assessment of the effectiveness of disturbance attenuation, are selected for the evaluation of the resultant effectiveness of the proposed approach. The resultant effectiveness of the proposed approach is compared with other selected speech-enhancement methods or denoising techniques available in the literature. The experimental research results show that the proposed method performs better than conventional methods in many types of high-noise conditions in terms of producing less residual noise and lower speech distortion.

DCT based densely connected convolutional GRU for real-time speech enhancement

Over the past ten years, deep learning has enabled significant advancements in the improvement of noisy speech. Due to the short time stability of speech signal, previous speech enhancement (SE) methods concentrated only on magnitude estimation, and these methods added a phase of the mixture in reconstructing the speech. The performance is limited in these approaches since the phase will also carry some of the speech information. Some of the speech enhancement approaches were developed later to jointly estimate both magnitudes as well as phases. Recently, complex-valued models, like deep complex convolution recurrent network (DCCRN), are proposed, but the computation of the model is very huge. In this work, we propose a Discrete Cosine Transform-based Densely Connected Convolutional Gated Recurrent Unit (DCTDCCGRU) model using dilated dense block and stacked GRU. The dense connectivity strengthens the gradient propagation by concatenating features from previous layers at the input. The advantage of the dense block is that at various resolutions, the dilated convolutions aid with context aggregation, and the dense connectivity provides a feature map with more precise target information by passing through multiple layers. To represent the correlation between neighboring noisy speech frames, a two Layer GRU is added in the bottleneck of U-Net. The experimental findings demonstrate that the proposed model outperformed the other existing models in terms of STOI (short-time objective intelligibility), PESQ (perceptual evaluation of the speech quality), and output SNR (signal-to-noise ratio).

Zero-sample mural superresolution reconstruction for enhanced perceptual quality

Aiming at the problem of texture loss and poor perceptual quality in low-resolution mural images, this paper proposes a zero-sample mural superresolution reconstruction method called EPZSSR to enhance perceptual quality, and the specific model is obtained by training the image. The algorithm takes the zero-shot superresolution method as the framework, randomly cuts the original image into a 128 * 128 size, performs Gaussian blurring on the image, uses Lanczos interpolation to downsample the smooth image to reduce artifacts, and uses convolutional attention module and skip connection to optimize the network structure. SmoothL1Loss is used to enhance the robustness of the model, and the PI value is introduced as the perceptual quality evaluation index. The experimental results show that compared with other superresolution reconstruction algorithms, the peak signal-to-noise ratio of the algorithm in this paper is increased by 0.98–3.23 dB on average. The mural texture reconstruction effect is better, the PI value is reduced by 0.56 on average, the mural perception quality is better, and the running time is reduced by 89.68 s on average. It has a certain value for mural superresolution reconstruction.

Shuffle Attention U-Net for Speech Enhancement in Time Domain

Over the past 10 years, deep learning has enabled significant advancements in the improvement of noisy speech. In an end-to-end speech enhancement, the deep neural networks transform a noisy speech signal to a clean speech signal in the time domain directly without any conversion or estimation of mask. Recently, the U-Net-based models achieved good enhancement performance. Despite this, some of them may neglect context-related information and detailed features of input speech in case of ordinary convolution. To address the above issues, recent studies have upgraded the performance of the model by adding various network modules such as attention mechanisms, long and short-term memory (LSTM). In this work, we propose a new U-Net-based speech enhancement model using a novel lightweight and efficient Shuffle Attention (SA), Gated Recurrent Unit (GRU), residual blocks with dilated convolutions. Residual block will be followed by a multi-scale convolution block (MSCB). The proposed hybrid structure enables the temporal context aggregation in time domain. The advantage of shuffle attention mechanism is that the channel and spatial attention are carried out simultaneously for each sub-feature in order to prevent potential noises while also highlighting the proper semantic feature areas by combining the same features from all locations. MSCB is employed for extracting rich temporal features. To represent the correlation between neighboring noisy speech frames, a two Layer GRU is added in the bottleneck of U-Net. The experimental findings demonstrate that the proposed model outperformed the other existing models in terms of short-time objective intelligibility (STOI), and perceptual evaluation of the speech quality (PESQ).

The Impact of Surgical Technique and Extent of Thyroidectomy in Voice Outcomes a Prospective Study

This prospective study aimed to evaluate the impact of surgical technique and extent of thyroidectomy on voice outcomes in 46 patients undergoing thyroidectomy. Patients were divided into three groups: limited thyroidectomy (n=20), total thyroidectomy (n=19), and extended thyroidectomy (n=7). Voice assessments were performed preoperatively and at 3 and 6 months postoperatively. Results showed that the surgical technique and extent of thyroidectomy had no significant effect on voice outcomes. However, a greater proportion of patients with limited thyroidectomy experienced a decrease in voice quality at three months postoperatively compared to patients with total or extended thyroidectomy. This study suggests that surgical technique and extent of thyroidectomy do not significantly affect voice outcomes in the long term. Further research is needed to understand these factors' impact on voice outcomes. Methodology: This prospective study was conducted in the department of surgery hmc Peshawar from Jan 2020 to Jan 2021 over the 01 years of a pride sample size of 46 patients (30 male and 16 female) who underwent thyroidectomy. Patients were divided into three groups based on the surgical technique and extent of thyroidectomy: limited thyroidectomy (n=20), total thyroidectomy (n=19), and extended thyroidectomy (n=7). All patients underwent preoperative and postoperative voice assessments at 3 and 6 months. The assessments included acoustic analysis and perceptual evaluation of voice quality. Results were compared between the three groups to evaluate the surgical technique's impact and thyroidectomy's extent on voice outcomes. Statistical analysis was performed to analyze the data. Results: The results showed that the surgical technique and extent of thyroidectomy had no significant effect on voice outcomes. However, a greater proportion of patients with limited thyroidectomy experienced a decrease in voice quality at three months postoperatively, compared to patients with total or extended thyroidectomy results shown in tables 1 to 7. Conclusion: This study suggests that surgical technique and extent of thyroidectomy do not significantly affect voice outcomes in the long term. Further research is needed to understand these factors' impact on voice outcomes. keywords: thyroidectomy, voice outcomes, surgical technique, the extent of thyroidectomy

Comparison of Full Factorial and Optimal Experimental Design for Perceptual Evaluation of Audiovisual Quality

Comparison of Full Factorial and Optimal Experimental Design for Perceptual Evaluation of Audiovisual Quality

Clinical Correlate of Dysphonia Measures in Patients with Parkinson’s Disease

In patients with Parkinson's disease (PD), the impairment of neurological control is frequently observed to affect the normal functioning of the muscles involved in speech production. In this study we propose to analyze a speech corpus from subjects with and without PD, with the aim of quantifying speech alterations. The data corpus was built in 2019 in collaboration with public hospitals and universities of Argentina, it contains 55 audio samples from patients with PD and 71 subjects without PD. The analysis was conducted from four perspectives: a) perceptual speech therapy evaluation of voice quality; b) estimation of acoustic disturbance through the integrated perturbance index; c) otorhinolaryngological clinical evaluation to identify structural and functional alterations; and d) acoustic analysis of the sustained phonation signal of the vowel /a/. In the latter, 339 measures of dysphonia were estimated, which were then reduced to 11 parameters using statistical feature selection methods. These parameters achieved a performance greater than 80% in the automatic binary classification task (PD or non-PD). In the otorhinolaryngological clinical evaluation analysis, the clinical findings were classified into four categories based on the presence or absence of the alteration: gastroesophageal reflux; tremor during phonation and rest; alterations in glottic closure; and other alterations. We estimated the degree of correlation between the analyses from these four perspectives. The integrated perturbance index is more relevant than the perceptual analysis of the voice quality, however, no correlation is observed with the clinical evaluation. Nonparametric statistical tests were used to assess mean differences between the 11 measures of dysphonia and each of the four categories of clinical findings identified. Significant differences were observed for four of the 11 selected dysphonia measures. These results encourage us to continue our analysis of acoustic parameters. We are also beginning to build a new data corpus, in which other measures of interest will be added.

Convolutional Transformer based Local and Global Feature Learning for Speech Enhancement

Speech enhancement (SE) is an important method for improving speech quality and intelligibility in noisy environments where received speech is severely distorted by noise. An efficient speech enhancement system relies on accurately modelling the long-term dependencies of noisy speech. Deep learning has greatly benefited by the use of transformers where long-term dependencies can be modelled more efficiently with multi-head attention (MHA) by using sequence similarity. Transformers frequently outperform recurrent neural network (RNN) and convolutional neural network (CNN) models in many tasks while utilizing parallel processing. In this paper we proposed a two-stage convolutional transformer for speech enhancement in time domain. The transformer considers global information as well as parallel computing, resulting in a reduction of long-term noise. In the proposed work unlike two-stage transformer neural network (TSTNN) different transformer structures for intra and inter transformers are used for extracting the local as well as global features of noisy speech. Moreover, a CNN module is added to the transformer so that short-term noise can be reduced more effectively, based on the ability of CNN to extract local information. The experimental findings demonstrate that the proposed model outperformed the other existing models in terms of STOI (short-time objective intelligibility), and PESQ (perceptual evaluation of the speech quality).

Acoustic Feedback Noise Cancellation in Hearing Aids Using Adaptive Filter

To enhance speech intelligibility for people with hearing loss, hearing aids will amplify speech using gains derived from evidence-based prescriptive methods, in addition to other advanced signal processing mechanisms. While the evidence supports the use of hearing aid signal processing for speech intelligibility, these signal processing adjustments can also be detrimental to hearing aid sound quality, with poor hearing aid sound quality cited as a barrier to device adoption. In general, an uncontrolled environment may contain degradation components like background noise, speech from other speakers etc. along with required speech components. In this paper, we implement adaptive filtering design for acoustic feedback noise cancellation in hearing aids. The adaptive filter architecture has been designed using normalized least mean square algorithm. By using adaptive filters both filter input coefficients are changeable during run-time and reduce noise in hearing aids. The proposed design is implemented in matlab and the simulations shows that the proposed architecture produces good quality of speech, accuracy, maintain stable steady state. The proposed design is validated with parameters like Noise Distortion, Perceptual Evaluation of Speech Quality, Signal to Noise Ratio, and Speech Distortion. The feedback canceller is implemented in MATLAB 9.4 simulink version release name of R2018a is used for validation with Echo Return Loss Enhancement (ERLE). The ERLE of the NMLS is reduced when the filter order is increases.Around 10% of the power spectrum density (PSD) is less when compared with existing designs.

Optimizing the Perceptual Quality of Time-Domain Speech Enhancement with Reinforcement Learning

In neural speech enhancement, a mismatch exists between the training objective, i.e., Mean-Square Error (MSE), and perceptual quality evaluation metrics, i.e., perceptual evaluation of speech quality and short-time objective intelligibility. We propose a novel reinforcement learning algorithm and network architecture, which incorporate a non-differentiable perceptual quality evaluation metric into the objective function using a dynamic filter module. Unlike the traditional dynamic filter implementation that directly generates a convolution kernel, we use a filter generation agent to predict the probability density function of a multivariate Gaussian distribution, from which we sample the convolution kernel. Experimental results show that the proposed reinforcement learning method clearly improves the perceptual quality over other supervised learning methods with the MSE objective function.

A speech enhancement algorithm based on a non-negative hidden Markov model and Kullback-Leibler divergence

In this paper, we propose a supervised single-channel speech enhancement method that combines Kullback-Leibler (KL) divergence-based non-negative matrix factorization (NMF) and a hidden Markov model (NMF-HMM). With the integration of the HMM, the temporal dynamics information of speech signals can be taken into account. This method includes a training stage and an enhancement stage. In the training stage, the sum of the Poisson distribution, leading to the KL divergence measure, is used as the observation model for each state of the HMM. This ensures that a computationally efficient multiplicative update can be used for the parameter update of this model. In the online enhancement stage, a novel minimum mean square error estimator is proposed for the NMF-HMM. This estimator can be implemented using parallel computing, reducing the time complexity. Moreover, compared to the traditional NMF-based speech enhancement methods, the experimental results show that our proposed algorithm improved the short-time objective intelligibility and perceptual evaluation of speech quality by 5% and 0.18, respectively.

Face hallucination using multisource references and cross‐scale dual residual fusion mechanism

There is an increasing interest in enhancing the quality of low-resolution (LR) facial images for various social life applications. Existing methods often use domain-specific prior knowledge, which is effective in improving the face super-resolution model's performance. However, it is challenging to obtain rich and accurate prior information from LR inputs in real-world scenarios, which can limit the robustness and generalization ability of the developed face super-resolution model. In this paper, a multisource reference-based face super-resolution Network, namely MSRNet, is proposed. Without considering the prior knowledge of faces, the network can reconstruct a LR face image with a magnitude factor of 8 under the guidance of multiple reference face images of different identities. By constructing an “appearance-alike” reference data set Face_Ref, the designed MSRNet aims to fully exploit the local and spatially similar high frequency information between the distinct references and the current face. More specifically, to effectively combine the information from multiple references, a cross-scale and cross-space feature fusion mechanism is introduced for external and internal references, and then the enhanced local semantics are finally incorporated into the high-resolution face reconstruction. The robustness of face image super-resolution is increased compared to current correlation approaches, since it not only eliminates the need for face prior knowledge but also avoids performing alignment operations on reference faces with multiple expressions and different poses. Experimental results show that the proposed model is able to produce results for face super-resolution that are satisfying and dependable and outperforms the state-of-the-art methods in terms of visual perceptual quality and quantity evaluation.

Validation of Acoustic Breathiness Index (ABI) in the South Indian Population

Acoustic Breathiness Index (ABI) is an acoustic model which equates to the auditory-perceptual (A-P) measure of breathiness. Globally, ABI has been validated in multiple languages. To date, empirical studies validating ABI in the South Indian population have not been carried out. Moreover, there is only minimal evidence on the impact of language and gender on ABI. This study attempted to validate ABI in languages of Malayalam and Kannada with "B" and "G" of GRBAS. The authors also examined the influence of language and gender on ABI. The authors retrospectively analyzed a dataset consisting of 160 voice samples (91 normophonic and 69 dysphonic). Five speech-language pathologists perceptually evaluated these samples on the GRBAS scale. PRAAT software was used to calculate ABI for each voice sample. The authors also determined the effects of language and gender on ABI. Following this, ABI's concurrent validity and diagnostic accuracy were examined relating to "B" and "G" of the GRBAS scale. The judges reached a moderate agreement on the perceptual evaluation of voice quality. The authors observed neither language nor gender differences in ABI. The concurrent validity of ABI was strong (r=0.761) when correlated against "B". Further, evaluations revealed that an ABI value of >3.66 could distinguish between breathy and non-breathy voice samples with an area under the curve of 0.816 (Sensitivity=61.73%; Specificity=94.93%). On the other hand, the concurrent validity of ABI in correlation with "G" did not yield the optimal result. ABI is a valid acoustic tool for identifying breathiness in the languages of Malayalam and Kannada. ABI correlates better with "B" than "G" on the GRBAS scale. ABI is independent of language and gender.

Gammatone Filter Bank-Deep Neural Network-based Monaural speech enhancement for unseen conditions

Speech signal enhancement achieves high-level performance in recent years using deep learning techniques. However, the deep learning technique in the speech enhancement algorithm degrades the performance of speech, particularly for unseen noises, unseen speakers and moreover, deep learning models are limited to the small number of speakers. Hence, we propose a Gammatone filterbank (GTFB) – simple deep neural network (SDNN) based speech enhancement algorithm to improve the quality of speech for three different unseen conditions. The use of GTFB gives a finer resolution in low-frequency regions of speech, and the SDNN model extracts a noisy GTFB frame as input and maps it to a clean speech GTFB frame. The experimental results are measured objectively using signal-noise-ratio, perceptual evaluation of speech quality, short time objective intelligibility, and subjectively using mean opinion score. The experimental results are carried out using a variety of training and testing models. The performance results show that the proposed GTFB-SDNN are robust to a variety of test situations and outperform existing methods.

Voice and Swallowing Disorders After Thyroid Surgery

The aim of the study was to evaluate voice and swallowing function following thyroid surgery, to identify dynamic changes during the follow-up period of 12-18 month and to find possible indicative signs of permanent or temporary vocal fold palsy. All participants (N=110) were prospectively enrolled from the preoperative thyroid surgery patients between September 2013 and December 2016. All subjects underwent preoperatively, first week and 12-18 month postoperatively videolaryngostroboscopy, filled in subjective evaluation of voice (voice handicap index, [VHI]) and swallowing (swallowing impairment score) complaints. Acoustic voice analysis (AVA), maximum phonation time (MPT) measurement and perceptual voice evaluation were conducted. In the presence of laryngeal damage, additional first and sixth-month follow-up visits were planned. Patients, whom we suspected laryngeal nerve damage, underwent laryngeal electromyography 4 weeks after the operation. On first postoperative week evaluation, no objective voice changes within patients with postoperatively intact laryngeal nerves were found. Subjective evaluations showed decline in VHI physical domain and increased strain in postoperative voice. Patients with postoperative nerve damage had a drastic decline in subjective voice quality (VHI all subscales and total score), AVA parameter jitter, MPT and perceptual voice quality. Subjective evaluation of swallowing revealed disturbances in all patients regardless of the nerve damage. We noticed remarkable differences in first week and first month visits between patients with permanent and temporary palsy in VHI total score and physical domain, MPT and perceptual breathiness and asthenia in voice. Regardless of the nerve injury, by the end of the follow-up period all changes had become statistically insignificant with the exception of perceptual evaluation of voice quality. Patients with postoperative laryngeal nerve damage experience substantial deterioration of both subjective and objective voice quality with more extensive impairment in patients with permanent paralysis. Thyroid surgery causes subjective swallowing changes irrespective of laryngeal nerve damage. In patients without laryngeal nerve damage, swallowing function improves following thyroidectomy. Possible indicators for permanent paralysis are delayed recovery in the values of MPT and jitter and persistent perceptual breathiness and asthenia.

Comparison and Validation of Acoustic Voice Quality Index Version 2 and Version 3 among South Indian Population.

Acoustic Voice Quality Index (AVQI) has emerged in the recent past as a robust multiparametric voice quality evaluation tool. Two versions of AVQI derived using the program PRAAT have found extensive clinical and research applications. These versions have been validated in several languages around the world. However, no research reports are available on validation of AVQI in the South Indian population. Further, studies comparing the performance of the two versions of AVQI are limited in the literature. This study was designed to validate and compare the two versions of AVQI (AVQIv02.02 and AVQIv03.01) in South Indian languages (Malayalam and Kannada). A retrospective analysis of previously recorded voice samples was carried out on a total of 160 (91 normophonic and 69 dysphonic) voice samples. These samples were perceptually rated on a GRBAS scale by five experienced speech-language pathologists. Standardized Syllable Number (SSN) necessary to derive AVQIv03.01 was computed. Following this, these samples were analyzed to obtain the AVQIv02.02 and AVQIv03.01. The concurrent validity and diagnostic accuracy of these measures were then examined and compared. A moderate agreement was obtained across the judges on perceptual evaluation of voice quality. SSN in Malayalam and Kannada languages were identified to be 29 and 25 syllables respectively. Language differences were not observed on both versions of AVQI. The concurrent validity of AVQIv03.01 (r=0.788) was superior to that of AVQIv02.02 (r=0.655). Further, the threshold of differentiating normophonic and dysphonic samples were determined to be >3.45 for AVQIv02.02 and >2.45 for AVQIv03.01. AVQIv03.01 is superior to AVQIv02.02 in terms of its diagnostic accuracy and concurrent validity. Current findings also extend the application of AVQI as a robust tool for the evaluation of voice characteristics to the South Indian population.