Peter van Hengel | University of Groningen (original) (raw)

Papers by Peter van Hengel

Objective. Automated speech recognition (ASR) systems have become increasingly sophisticated, acc... more Objective. Automated speech recognition (ASR) systems have become increasingly sophisticated, accurate, and deployable on many digital devices, including on a smartphone. This pilot study aims to examine the speech recognition performance of ASR apps using audiological speech tests. In addition, we compare ASR speech recognition performance to normal hearing and hearing impaired listeners and evaluate if standard clinical audiological tests are a meaningful and quick measure of the performance of ASR apps.Methods. Four apps have been tested on a smartphone, respectively AVA (iOS, Android), Earfy (iOS, Android), Live Transcribe (Android) and Speechy (iOS). The Dutch audiological speech tests performed were speech audiometry in quiet (Dutch CNC-test), Digits-in-Noise (DIN)-test with steady-state speech-shaped noise, sentences in quiet and in averaged long-term speech-shaped spectrum noise (Plomp-test). For comparison, the apps’ ability to transcribe a spoken dialogue (Dutch and Englis...

ORL, 2001

In this study we performed standardized measurements of spontaneous nystagmus (SN), caloric tests... more In this study we performed standardized measurements of spontaneous nystagmus (SN), caloric tests and velocity step tests on a well-defined group of 92 Ménière patients as part of a diagnostic protocol. The aim of the study was to look for results that could contribute to the diagnosis, but more specifically to a possible classification, of Ménière’s disease. Relations with average hearing loss, shape of the audiogram, average hearing loss, duration and severity of perceived symptoms of disease and differences between uni-and bilateral Ménière’s disease were studied. Directional preponderance (DP) in unilaterally affected patients and labyrinthine preponderance (LP) in both uni- and bilaterally affected patients were significantly more often directed towards the unaffected (uni) or least affected (bi) side (p = 0.007, p = 0.001, p = 0.002, respectively). DP was correlated with LP (uni: R = 0.615, p < 0.001; bi: R = 0.438, p < 0.01), which means that the DP can be seen as a lat...

Informatics for Health and Social Care, 2010

Worldwide, ageing societies are bringing challenges for independent living and healthcare. Health... more Worldwide, ageing societies are bringing challenges for independent living and healthcare. Health-enabling technologies for pervasive healthcare and sensor-enhanced health information systems offer new opportunities for care. In order to identify, implement and assess such new information and communication technologies (ICT) the &amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;#39;Lower Saxony Research Network Design of Environments for Ageing&amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;#39; (GAL) has been launched in 2008 as interdisciplinary research project. In this publication, we inform about the goals and structure of GAL, including first outcomes, as well as to discuss the potentials and possible barriers of such highly interdisciplinary research projects in the field of health-enabling technologies for pervasive healthcare. Although GAL&amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;#39;s high interdisciplinarity at the beginning slowed down the speed of research progress, we can now work on problems, which can hardly be solved by one or few disciplines alone. Interdisciplinary research projects on ICT in ageing societies are needed and recommended.

The newly founded project group Hearing, Speech and Audio Technology of the Fraunhofer Institute ... more The newly founded project group Hearing, Speech and Audio Technology of the Fraunhofer Institute for Digital Media Technology IDMT focuses on applications based on models of the human hearing. One of the domains for application of these models is the detection of incidents in the surveillance domain based on audio information. For this application a system designed to detect human verbal aggression was developed by Sound Intelligence. A case study with this system will be described here as an example of what can be achieved using audio technology in a supportive role for surveillance .Models of human sound processing incorporating aspects of neural processing are now further developed to indicate a surveillance operator possible incidents or unexpected events.

Modern hearing aids provide many parameters that can be adjusted to optimize the hearing experien... more Modern hearing aids provide many parameters that can be adjusted to optimize the hearing experience of the individual user. Optimization of these parameters can be based on a comparison of an internal representation of sound processed by the hearing aid and the impaired hearing system with the representation in a non-impaired ear. Models that can represent the most common types of hearing loss and can be adjusted to fit individual hearing loss can play a crucial role in such optimization procedures. Simulations are presented that show the potential of a transmission line model in such a procedure. The model is extended to remap cochleogram energy based on estimations of the local instantaneous frequency. This ‘remapping’ of the cochleogram gives an advantage in tone-in-noise detection that may be related to neural deafferentation.

Recent Developments in Auditory Mechanics, 2000

Hearing impairment is a growing problem for our ageing communication society. Although many elder... more Hearing impairment is a growing problem for our ageing communication society. Although many elderly people take notice of their hearing deficiencies in daily communication, only few of them are supplied with hearing aids. This maybe a cause of several psycho-social factors (e.g. stigmatization, comfort, cost and effort) associated with hearing loss . The Hearing at Home (HaH) project [1] tries to tackle this general acceptance problem by applying supportive hearing technology into (highly accepted) home-entertainment devices, allowing the elderly people to participate in communica-tion without using hearing aids. It is anticipated that formerly separated devices like TV, telephone, fax, intercom, per-sonal computer, HIFI systems as well as services like VoIP and home automation grow together and can be integrated into a single Home Information and Communication (HIC) platform accessible and controlled through a TV set. The HIC platform that is researched and developed in HaH will su...

Biophysics of the Cochlea, 2003

ABSTRACT We have developed a flexible 1D cochlea model to test hypotheses and data against physic... more ABSTRACT We have developed a flexible 1D cochlea model to test hypotheses and data against physical and mathematical constraints. The model is flexible in the sense that several linear and nonlinear model characteristics can be selected, and different boundary conditions can be tested. The software model runs at a reasonable speed at a modern PC. As an example, we will show the results of the model in comparison with the systematic study of the phase behavior (group delay) of distortion product otoacoustic emissions (DPOAEs) in the guinea pig (S. Schneider, V. Prijs and R. Schoonhoven, [9]). We also will demonstrate the effects of some common non-physical boundary conditions. Finally, we briefly indicate that this model of the auditory periphery provides a superior front end for an ASR (automatic speech recognition)-system.

Informatics for Health and Social Care, 2014

2007 IEEE Conference on Advanced Video and Signal Based Surveillance, 2007

The paper presents a knowledge-based system designed to detect evidence of aggression by means of... more The paper presents a knowledge-based system designed to detect evidence of aggression by means of audio analysis. The detection is based on the way sounds are analyzed and how they attract attention in the human auditory system. The performance achieved is comparable to human performance in complex social environments. The SIgard system has been deployed in a number of different

Audio, Speech, and Language Processing, IEEE/ACM Transactions on, 2014

ABSTRACT This article compares methods for the conversion of timeseries into a spectro-temporal r... more ABSTRACT This article compares methods for the conversion of timeseries into a spectro-temporal representation. These methods are designed based on a resemblance with the auditory processing of sound in the mammalian inner ear, or on mathematical principles related to, for example, Fourier analysis. This study provides a comparison between several of these methods. Two tests were devised for this comparison: one based on susceptibility to noise and one on the expression of spectrotemporal detail. These two aspects were considered of importance for real world applications. While some methods produced good results on one of the two tests, others produced good results on both. Overall the transmission line model using an impedance function suggested by Zweig [1] provided the best results, though not significantly. Also a larger computational load may hinder application in some domains. The gammatone filterbank and straightforward spectrogram provide good alternatives with less computational load. The introduction of nonlinearity was shown to deteriorate performance on both tests, in both the filterbank and in the transmission line model.

Ambient Assisted Living, 2011

ABSTRACT Due to the demographic change, the number of older people will grow steadily within the ... more ABSTRACT Due to the demographic change, the number of older people will grow steadily within the next decades [1]. Technical systems, methods and algorithms developed in the context of Ambient Assisted Living (AAL) aim to enable these people to live self-determined and safe in their own homes as long as possible. To ensure this, these homes may have to be equipped with e.g., a health monitoring system, special sensors or devices for gathering and evaluating individual information. In this paper, an acoustic monitoring system for evaluation of acoustic cues is presented and tested exemplarily on four acoustic events (cough, knock, clap and phone bell). The acoustic monitoring includes several consecutive signal processing steps: the audio signal picked up by a microphone is transformed in a psycho-physiologically weighted domain using a gammatone filterbank, resulting in a so-called cochleogram [2]. From this cochleogram, background noise is estimated and eliminated. Further signal processing determines partitions of the cochleogram which are considered to form acoustic events. Each of these events is then evaluated with a classification algorithm. In this classification step, class membership is determined by comparing the event under inspection to representatives which have been calculated during a training phase utilizing a self-organizing map [3]. It is shown that a true positive rate from over 79% can be achieved where the false positive rate is smaller than 4% except for the event knocking.

Otology & Neurotology, 2002

In this study, three-dimensional Fourier transformation constructive interference in steady state... more In this study, three-dimensional Fourier transformation constructive interference in steady state (3DFT-CISS) magnetic resonance imaging was used to quantify the distance between the vertical part of the posterior semicircular canal and the posterior fossa as a measure of the endolymphatic sac and duct in patients with Menière&amp;#39;s disease. Differences in this distance between affected and unaffected ears, as well as differences between unilaterally and bilaterally affected patients, were studied and compared with a control group. Also, possible correlations between the measured distance and the duration and severity of symptoms, patient age, and average hearing loss were investigated in the group of patients with Menière&amp;#39;s disease. Retrospective clinical study. Tertiary referral center (University Hospital) as part of a large, diagnostic research project on Menière&amp;#39;s disease. Of the 111 patients with Menière&amp;#39;s disease initially included, 90 patients underwent 3DFT-CISS MRI. Eighty-six of these patients were analyzed in this MRI study. Fifty-six patients had unilateral Menière&amp;#39;s disease, and 30 patients had bilateral Menière&amp;#39;s disease (116 affected and 56 unaffected ears). Sixty-two ears in patients without Menière&amp;#39;s disease were studied as controls. The distance between the vertical part of the posterior semicircular canal and the posterior fossa was determined by 3DFT-CISS MRI. Contiguous axial 3DFT-CISS MRI slices of 0.7 to 1.0 mm were made by a radiologist according to a strict protocol. Measurements of the distance between the vertical part of the posterior semicircular canal and the posterior fossa were taken by two professionals-a radiologist and an otolaryngologist-using a ruler and the original scan. A significantly smaller distance (2.9 mm) between the vertical part of the posterior semicircular canal and the posterior fossa as visualized on MRI scans was found in the ears of patients with Menière&amp;#39;s disease than in the ears of patients in the control group (3.8 mm, p &amp;lt; 0.001). In both uni- and bilaterally affected patients (n = 56 and n = 30, respectively), no significant difference between ears was found (p = 0.44 and p = 0.19, respectively). In bilaterally affected patients, however, this distance (3.2 mm) was significantly greater than the distance in unilaterally affected patients (2.7 mm, p = 0.004). There was no relationship between the MRI-visualized distance between the vertical part of the posterior semicircular canal and the posterior fossa and the duration of disease, average hearing loss, or severity of symptoms in uni- and bilaterally affected patients. The difference in MRI-visualized distances between the vertical part of the posterior semicircular canal and the posterior fossa of uni- and bilaterally affected patients strongly suggests that unilateral and bilateral hearing loss are two different entities in patients with Menière&amp;#39;s disease. The size of the endolymphatic sac seems not to be the only factor in the pathogenesis of Menière&amp;#39;s disease. That the MRI-visualized distance between the vertical part of the posterior semicircular canal and the posterior fossa does not have any relationship to the duration of the disease or to patient age indicates that this distance is a congenital feature.

Clinical Otolaryngology and Allied Sciences, 2001

Advanced Video and Signal Based Surveillance, 2007

The Journal of the …, 1999

Distortion product otoacoustic emissions ͑DPOAE͒ were recorded from eight human subjects with mil... more Distortion product otoacoustic emissions ͑DPOAE͒ were recorded from eight human subjects with mild to moderate cochlear hearing loss, using a frequency spacing of 48 primary pairs per octave and at a level L 1 ϭL 2 ϭ60 dBSPL and with a fixed ratio f 2 / f 1. Subjects with different shapes of hearing thresholds were selected. They included subjects with near-normal hearing within only a limited frequency range, subjects with a notch in the audiogram, and subjects with a mild to moderate high-frequency loss. If the primaries were located in a region of normal or near-normal hearing, but DP frequencies were located in a region of raised thresholds, the distortion product 2 f 1-f 2 was still observable, but the DP fine structure disappeared. If the DP frequencies fell into a region of normal thresholds, fine structure was preserved as long as DPOAE were generated, even in cases of mild hearing loss in the region of the primaries. These experimental results give further strong evidence that, in addition to the initial source in the primary region, there is a second source at the characteristic place of f DP. Simulations in a nonlinear and active computer model for DPOAE generation indicate different generation mechanisms for the two components. The disappearance of DPOAE fine structure might serve as a more sensitive indicator of hearing impairment than the consideration of DP level alone.