Speech quality evaluation of a sparse coding shrinkage noise reduction algorithm with normal hearing and hearing impaired listeners

Sang, Jinqiu; Hu, Hongmei; Zheng, Chengshi; Li, Guoping; Lutman, Mark E.; Bleeck, Stefan

doi:10.1016/j.heares.2015.07.019

Cited by 13 publications

(6 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The recognition ability of NH and hearing-impaired (HI) listeners 10 is increased using the "auditory masked threshold noise suppression" technique 10 under noisy conditions. 11 Conversely, babble noise for HI listeners, 12 speech-shaped quality and speech intelligibility in speech-shaped noise were improved using the laboratory-tested sparse code shrinkage algorithm. Past works have reported that using single-channel enhancement algorithms for HI listeners has shown poor performance in word recognition but not for listener preference.…”

Section: Introductionmentioning

confidence: 99%

RETRACTED: Speech enhancement method using deep learning approach for hearing-impaired listeners

Rahiman

Jayanthi

Jayanthi³

2020

Health Informatics J

View full text Add to dashboard Cite

A deep learning-based speech enhancement method is proposed to aid hearing-impaired listeners by improving speech intelligibility. The algorithm decomposes the noisy speech signal into frames (as features). Subsequently, a deep convolutional neural network is fed with decomposed noisy speech signal frames to produce frequency channel estimation. However, a higher signal-to-noise ratio information is contained in produced frequency channel estimation. Using this estimate, speech-dominated cochlear implant channels are taken to produce electrical stimulation. This process is the same as that of the conventional n-of- m cochlear implant coding strategies. To determine the speech-in-noise performance of 12 cochlear implant users, the fan and music sound applied are considered as background noises. Performance of the proposed algorithm is evaluated by considering these background noises. Low processing delay and reliable architecture are the best characteristics of the deep learning-based speech enhancement algorithm; hence, this can be suitably applied for all applications of hearing devices. Experimental results demonstrate that deep convolutional neural network approach appeared more promising than conventional approaches.

show abstract

Section: Introductionmentioning

confidence: 99%

RETRACTED: Speech enhancement method using deep learning approach for hearing-impaired listeners

Rahiman

Jayanthi

Jayanthi³

2020

Health Informatics J

View full text Add to dashboard Cite

show abstract

“…At frequencies of 2.0–8.0 kHz (especially 3.0–4.5 kHz), the human ear has a higher sensitivity (the Fletcher–Munson curves of equal volume levels ISO 226: 2003), and sounds are perceived as being 10–20 dB louder than those outside this range, at same intensity [21]. Furthermore, in this frequency range, essential parts of speech information are located [22], impeding communication within the OT team. Persistent, high levels of noise are known to lead to health problems [23–26].…”

Section: Discussionmentioning

confidence: 99%

Modifications of Surgical Suction Tip Geometry for Flow Optimisation: Influence on Suction-Induced Noise Pollution

Friedrich

Tirilomis

Kollmeier

et al. 2018

Surgery Research and Practice

View full text Add to dashboard Cite

Introduction Suction devices for clearing the surgical field are among the most commonly used tools of every surgeon because a better view of the surgical field is essential. Forced suction may produce disturbingly loud noise, which acts as a nonnegligible stressor. Especially, in emergency situations with heavy bleeding, this loud noise has been described as an impeding factor in the medical decision-making process. In addition, there are reports of inner ear damage in patients due to suction noises during operations in the head area. These problems have not been solved yet. The purpose of this study was to analyse flow-dependent suction noise effects of different surgical suction tips. Furthermore, we developed design improvements to these devices. Methods We compared five different geometries of suction tips using an in vitro standardised setup. Two commercially available standard suction tips were compared to three adapted new devices regarding their flow-dependent (10–2000 mL/min) noise emission (dB, weighting filter (A), distance 10 cm) and acoustic quality of resulting noises (Hamilton fast Fourier analysis) during active suction at the liquid-air boundary. Noise maps at different flow rates were created for all five suction devices, and the proportion of extracted air was measured. The geometries of the three custom-made suction tips (new models 1, 2, and 3) were designed considering the insights after determining the key characteristics of the two standard suction models. Results The geometry of a suction device tip has significant impact on its noise emission. For the standard models, the frequency spectrum at higher flow rates significantly changes to high-frequency noise patterns (>3 kHz). A number of small side holes designed to prevent tissue adhesion lead to increased levels of high-frequency noise. Due to modifications of the tip geometry in our new models, we are able to achieve a highly significant reduction of noise level at low flow rates (new model 2 vs. standard models p < 0.001) and also the acoustic quality improved. Additionally, we attain a highly significant reduction of secondary air intake (new model 2 vs. the other models p < 0.001). Conclusion Improving flow-relevant features of the geometry of suction heads is a suitable way to reduce noise emissions. Optimized suction tips are significantly quieter. This may help us to reduce noise-induced hearing damage in patients as well as stress of medical staff during surgery and should lead to quieter operation theatres overall. Furthermore, the turbulence reduction and reduced secondary air intake during the suction process are expected to result in protective effects on the collected blood and thus could improve the quality of autologous blood retransfusions. We are on the way to evaluate potential benefits.

show abstract

“…Increasing sound level more than 12 dB over the background noise does not affect significant increase in speech quality 26. Despite optimal volume and frequency transmission, speech quality and definition decreases with increased reverberation times.…”

Section: Discussionmentioning

confidence: 87%

New technical solution to minimise noise exposure for surgical staff: the ‘silent operating theatre optimisation system’

et al. 2017

View full text Add to dashboard Cite

The increasing number of technical equipment in the operating room (OR) is resulting in significantly higher noise levels. The more complex and sophisticated the surgical procedure is, the more essential it is for all team members of the OR to work together in a harmonious fashion to process and manage their demanding team tasks. With increasing noise in OR, the risk of more frequent errors also increases. The reduction of noise production in the OR is possible but limited. The aim of this study was to develop a device that reduces ambient noise for the operating team without hindering their ability to communicate.We developed a new communication technology set-up for the OR to meet all needed requirements. All members of the operating team are issued headsets with microphones. The headsets filter out background noises (active and passive noise cancelling) and the microphones enable interactive communication among and between OR subgroups through targeted information selection (signal selection). Any remaining background noise is overshadowed by music, which is quiesced by direct speech into the microphone (ducking). Information flow is programmed on a digital workstation, providing each team member a selection of acoustic signals from the OR on their bidirectional headset. A complex matrix of connections in this audio technology allows a predefined communication structure. These procedures were assembled in the Silent Operating Theatre Optimisation System (SOTOS). The technical specifications and user interface are described. A pilot study in 2015 using the SOTOS in cardiac surgery showed very positive feedback from the participating operating team members. Further studies focusing on communicational psychology perspective and physiological reaction are recommended.

show abstract

Speech quality evaluation of a sparse coding shrinkage noise reduction algorithm with normal hearing and hearing impaired listeners

Cited by 13 publications

References 25 publications

RETRACTED: Speech enhancement method using deep learning approach for hearing-impaired listeners

RETRACTED: Speech enhancement method using deep learning approach for hearing-impaired listeners

Modifications of Surgical Suction Tip Geometry for Flow Optimisation: Influence on Suction-Induced Noise Pollution

New technical solution to minimise noise exposure for surgical staff: the ‘silent operating theatre optimisation system’

Contact Info

Product

Resources

About