Biomarker-Informed Machine Learning Model of Cognitive Fatigue from a Heart Rate Response Perspective.

Cognitive fatigue is a psychological state characterised by feelings of tiredness and impaired cognitive functioning arising from high cognitive demands. This paper examines the recent research progress on the assessment of cognitive fatigue and provides informed recommendations for future research. Traditionally, cognitive fatigue is introspectively assessed through self-report or objectively inferred from a decline in behavioural performance. However, more recently, researchers have attempted to explore the biological underpinnings of cognitive fatigue to understand and measure this phenomenon. In particular, there is evidence indicating that the imbalance between sympathetic and parasympathetic nervous activity appears to be a physiological correlate of cognitive fatigue. This imbalance has been indexed through various heart rate variability indices that have also been proposed as putative biomarkers of cognitive fatigue. Moreover, in contrast to traditional inferential methods, there is also a growing research interest in using data-driven approaches to assessing cognitive fatigue. The ubiquity of wearables with the capability to collect large amounts of physiological data appears to be a major facilitator in the growth of data-driven research in this area. Preliminary findings indicate that such large datasets can be used to accurately predict cognitive fatigue through various machine learning approaches. Overall, the potential of combining domain-specific knowledge gained from biomarker research with machine learning approaches should be further explored to build more robust predictive models of cognitive fatigue.

Using empirical wavelet transform to speed up selective filtered active noise control system.

The gradual adaptation and possibility of divergence hinder the active noise control system from being applied to a wider range of applications. Selective active noise control has been proposed to rapidly reduce noise by selecting a pre-trained control filter for different primary noise detected without an error microphone. For stationary noise, considerable noise reduction performance with a short selection period is obtained. For non-stationary noise, more restrictive requirements are imposed on instant convergence, as it leads to faster tracking and better noise reduction performance. To speed up a selective filtered active noise control system, empirical wavelet transform is introduced here to accurately and instantaneously extract the frequency information of primary noise. The boundary of the first intrinsic mode function of random noises is extracted as the instant signal feature. Primary noise is attenuated immediately by picking the optimal pre-trained control filter labeled by the nearest boundary. The storage requirement for a pre-trained control filter library is reduced. Instant control is obtained, and the instability caused by output saturation is overcome. With more concentrated energy distribution, better noise reduction performance is achieved by the proposed algorithm compared to conventional and selective active noise control algorithms. Simulation results validate these advantages of the proposed algorithm.

Feedforward multichannel virtual-sensing active control of noise through an aperture: Analysis on causality and sensor-actuator constraints.

The multichannel implementation of the auxiliary-filter-based virtual-sensing (AF-VS) technique for active noise control applications is revisited and realized in the paper. Frequency-domain analysis based on random primary noise proves that the multichannel virtual-sensing active noise control (MVANC) technique can achieve optimal control at the desired virtual locations even if the signals at the physical and virtual microphones are not causally related. Further analysis on a number of sensor-actuator configurations shows that the MVANC technique achieves optimal control at the desired locations as long as the number of secondary sources does not exceed that of the physical error microphones. Furthermore, the simulations with measured transfer functions and real-time experiments conducted on a four-channel system validate the frequency domain analyses.

Modeling distance-dependent individual head-related transfer functions in the horizontal plane using frontal projection headphones.

The veracity of virtual audio is degraded by the use of non-individualized head-related transfer functions (HRTFs) due to the introduction of front-back, elevation confusions, and timbral coloration. Hence, an accurate reproduction of spatial sound demands the use of individualized HRTFs. Measuring distance-dependent individualized HRTFs can be extremely tedious, since it requires precise measurements at several distances in the proximal region (<1 m) for each individual. This paper proposes a technique to model distance-dependent individualized HRTFs in the horizontal plane using "frontal projection headphones playback" that does not require individualized measurements. The frontal projection headphones [Sunder, Tan, and Gan (2013). J. Audio Eng. Soc. 61, 989-1000] project the sound directly onto the pinnae from the front, and thus inherently create listener's idiosyncratic pinna cues at the eardrum. Perceptual experiments were conducted to investigate cues (auditory parallax and interaural level differences) that aid distance perception in anechoic conditions. Interaural level differences were identified as the prominent cue for distance perception and a spherical head model was used to model these distance-dependent features. Detailed psychophysical experiments revealed that the modeled distance-dependent individualized HRTFs exhibited localization performance close to the measured distance-dependent individualized HRTFs for all subjects.

What is behind the meta-learning initialization of adaptive filter? - A naive method for accelerating convergence of adaptive multichannel active noise control.

Active noise control (ANC) is a typical signal-processing technique that has recently been utilized extensively to combat the urban noise problem. Although numerous advanced adaptive algorithms have been devised to enhance noise reduction performance, few of them have been implemented in actual ANC products due to their high computational complexity and slow convergence. With the rapid development of deep learning technology, Meta-learning-based initialization appears to become an efficient and cost-effective method for accelerating the convergence of adaptive algorithms. However, few dedicated Meta-learning algorithms exist for adaptive signal processing applications, particularly multichannel active noise control (MCANC). Hence, we proposed a modified Model-Agnostic Meta-Learning (MAML) initialization for the MCANC system.1 Additional theatrical research reveals that the nature of MAML, when applied to signal processing, is the expectation of a weight-sum gradient. Based on this discovery, we devised the Monte-Carlo Gradient Meta-learning (MCGM) algorithm, which employed a more straightforward procedure to accomplish the same performance as the Modified MAML algorithm. Furthermore, the numerical simulation of ANC using raw noise samples on measured paths validates the efficacy of the proposed methods in accelerating the convergence of the multichannel-filtered reference least mean square algorithm (McFxLMS).

Lion city soundscapes: Modified partitioning around medoids for a perceptually diverse dataset of Singaporean soundscapesa).

This study presents a dataset of audio-visual soundscape recordings at 62 different locations in Singapore, initially made as full-length recordings over spans of 9-38 min. For consistency and reduction in listener fatigue in future subjective studies, one-minute excerpts were cropped from the full-length recordings. An automated method using pre-trained models for Pleasantness and Eventfulness (according to ISO 12913) in a modified partitioning around medoids algorithm was employed to generate the set of excerpts by balancing the need to encompass the perceptual space with uniformity in distribution. A validation study on the method confirmed its adherence to the intended design.

Product directivity models for parametric loudspeakers.

In a recent work, the beamsteering characteristics of parametric loudspeakers were validated in an experiment. It was shown that based on the product directivity model, the locations and amplitudes of the mainlobe and grating lobes could be predicted within acceptable errors. However, the measured amplitudes of sidelobes have not been able to match the theoretical results accurately. In this paper, the original theories behind the product directivity model are revisited, and three modified product directivity models are proposed: (i) the advanced product directivity model, (ii) the exponential product directivity model, and (iii) the combined product directivity model. The proposed product directivity models take the radii of equivalent Gaussian sources into account and obtain better predictions of sidelobes for the difference frequency waves. From the comparison between measurement results and numerical solutions, all the proposed models outperform the original product directivity model in terms of selected sidelobe predictions by about 10 dB.

Lowering the Sampling Rate: Heart Rate Response during Cognitive Fatigue.

Cognitive fatigue is a mental state characterised by feelings of tiredness and impaired cognitive functioning due to sustained cognitive demands. Frequency-domain heart rate variability (HRV) features have been found to vary as a function of cognitive fatigue. However, it has yet to be determined whether HRV features derived from electrocardiogram data with a low sampling rate would remain sensitive to cognitive fatigue. Bridging this research gap is important as it has substantial implications for designing more energy-efficient and less memory-hungry wearables to monitor cognitive fatigue. This study aimed to examine (1) the level of agreement between frequency-domain HRV features derived from lower and higher sampling rates, and (2) whether frequency-domain HRV features derived from lower sampling rates could predict cognitive fatigue. Participants (N = 53) were put through a cognitively fatiguing 2-back task for 20 min whilst their electrocardiograms were recorded. Results revealed that frequency-domain HRV features derived from sampling rate as low as 125 Hz remained almost perfectly in agreement with features derived from the original sampling rate at 2000 Hz. Furthermore, frequency domain features, such as normalised low-frequency power, normalised high-frequency power, and the ratio of low- to high-frequency power varied as a function of increasing cognitive fatigue during the task across all sampling rates. In conclusion, it appears that sampling at 125 Hz is more than adequate for frequency-domain feature extraction to index cognitive fatigue. These findings have significant implications for the design of low-cost wearables for detecting cognitive fatigue.

Automation of binaural headphone audio calibration on an artificial head.

In studies with auralisation of audio stimuli over headphones, accurate presentation of headphone audio is critical for replicability and ecological validity. Audio stimuli levels are usually calibrated by placing studio quality headphones on an artificial head and torso simulator. Manual adjustment of audio tracks becomes laborious when the number of stimuli is large, especially for applications with large datasets. To increase reliability and productivity, we devised a stimulus-agnostic, automated calibration procedure for headphone audio via an artificial head and torso simulator, with a LabVIEW implementation available at doi:10.21979/N9/0KYIAU.•The procedure uses a National Instruments NI-9234 data acquisition module and works with any ITU­T P.58:2013 and ANSI/ASA S 3.36:2012 compliant artificial head measurement systems.•The procedure works by an adjustment to a generic guess, followed by a modified binary search, wherein the audio stimuli are calibrated to within a user-specified tolerance level.•Each stimulus in a validation run to calibrate 250 stimuli to 65.0 ± 0.5 dB was played back an average of 2.22 ± 0.92 times before successful calibration, thus demonstrating the robustness and efficiency of our proposed method.

Active control of broadband sound through the open aperture of a full-sized domestic window.

Shutting the window is usually the last resort in mitigating environmental noise, at the expense of natural ventilation. We describe an active sound control system fitted onto the opening of the domestic window that attenuates the incident sound, achieving a global reduction in the room interior while maintaining natural ventilation. The incident sound is actively attenuated by an array of control modules (a small loudspeaker) distributed optimally across the aperture. A single reference microphone provides advance information for the controller to compute the anti-noise signal input to the loudspeakers in real-time. A numerical analysis revealed that the maximum active attenuation potential outperforms the perfect acoustic insulation provided by a fully shut single-glazed window in ideal conditions. To determine the real-world performance of such an active control system, an experimental system is realized in the aperture of a full-sized window installed on a mockup room. Up to 10-dB reduction in energy-averaged sound pressure level was achieved by the active control system in the presence of a recorded real-world broadband noise. However, attenuation in the low-frequency range and its maximum power output is limited by the size of the loudspeakers.

Effects of adding natural sounds to urban noises on the perceived loudness of noise and soundscape quality.

Introducing pleasant natural sounds to mask urban noises is an important soundscape design strategy to improve acoustic comfort. This study investigates the effects of signal-to-noise ratio (SNR) between natural sounds (signal) and the target noises (noise) and their temporal characteristics on the perceived loudness of noise (PLN) and overall soundscape quality (OSQ) through a laboratory experiment. Two types of urban noise sources (hydraulic breaker and traffic noises) were set to A-weighted equivalent sound pressure levels (SPL) of 55, 65, and 75 dB and then augmented with two types of natural sounds (birdsong and stream), across a range of SNRs. Each acoustic stimulus was a combination of noise and natural sound at SNRs from -6 to 6 dB. Averaged across all cases, the subjective assessment of PLN showed that augmenting urban noise separately with the two natural sounds reduced the PLN by 17.9%, with no significant differences found between the birdsong and stream sounds. Adding natural sounds increased the OSQ by on average 18.3% across the cases, but their effects gradually decreased as the noise level increased. The OSQ of the birdsong and stream sounds were similar for traffic noise, whereas the stream sound was rated higher than the birdsong for the breaker noise. The results suggest that increasing the dissimilarity in temporal structure between the target noise and natural sounds could enhance the soundscape quality. Appropriate SNRs were explored considering both PLN and OSQ. The results showed that the SNR of -6 dB was desirable when the A-weighted SPL of the noise rose to 75 dB.

Modeling of finite-amplitude sound beams: second order fields generated by a parametric loudspeaker.

The nonlinear interaction of sound waves in air has been applied to sound reproduction for audio applications. A directional audible sound can be generated by amplitude-modulating the ultrasound carrier with an audio signal, then transmitting it from a parametric loudspeaker. This brings the need of a computationally efficient model to describe the propagation of finite-amplitude sound beams for the system design and optimization. A quasilinear analytical solution capable of fast numerical evaluation is presented for the second-order fields of the sum-, difference-frequency and second harmonic components. It is based on a virtual-complex-source approach, wherein the source field is treated as an aggregation of a set of complex virtual sources located in complex distance, then the corresponding fundamental sound field is reduced to the computation of sums of simple functions by exploiting the integrability of Gaussian functions. By this result, the five-dimensional integral expressions for the second-order sound fields are simplified to one-dimensional integrals. Furthermore, a substantial analytical reduction to sums of single integrals also is derived for an arbitrary source distribution when the basis functions are expressible as a sum of products of trigonometric functions. The validity of the proposed method is confirmed by a comparison of numerical results with experimental data previously published for the rectangular ultrasonic transducer.

A complex virtual source approach for calculating the diffraction beam field generated by a rectangular planar source.

In this paper, a complex virtual source approach for calculating the ultrasound field generated by a rectangular planar source is presented. Instead of using a real rectangular plane source, the equivalent sources that have complex amplitudes in complex space are used to compute the sound field distribution. The parabolic equation first is solved in the kappa-space domain by applying Fourier transform. The kappa-space domain source is then expressed as a set of Gaussian functions, and the related coefficients is determined by the optimization method. The analytic solution then is derived, and the effect of the parameters on the calculation accuracy is discussed. The comparison between the proposed fast numerical scheme and previous methods (Fresnel integral and Ocheltree's method) and are given in an example. The numerical results reveal that the computation time in obtaining accurate calculations is greatly reduced by using the proposed method.

Grating lobe elimination in steerable parametric loudspeaker.

In the past two decades, the majority of research on the parametric loudspeaker has concentrated on the nonlinear modeling of acoustic propagation and pre-processing techniques to reduce nonlinear distortion in sound reproduction. There are, however, very few studies on directivity control of the parametric loudspeaker. In this paper, we propose an equivalent circular Gaussian source array that approximates the directivity characteristics of the linear ultrasonic transducer array. By using this approximation, the directivity of the sound beam from the parametric loudspeaker can be predicted by the product directivity principle. New theoretical results, which are verified through measurements, are presented to show the effectiveness of the delay-and-sum beamsteering structure for the parametric loudspeaker. Unlike the conventional loudspeaker array, where the spacing between array elements must be less than half the wavelength to avoid spatial aliasing, the parametric loudspeaker can take advantage of grating lobe elimination to extend the spacing of ultrasonic transducer array to more than 1.5 wavelengths in a typical application.

The DORT solution in acoustic inverse scattering problem of a small elastic scatterer.

The DORT (French acronym for Décomposition de l'Opérateur de Retournement Temporel) method is a novel approach for active detection and focusing of acoustic waves on the targets in the scattering medium. This technique involves the determination of the invariant of the time-reversal operator obtained by measurement of the scattering data in a pulse-echo mode. In this paper, a proposed approach based on the DORT method is developed to solve the acoustic inverse scattering problem of a small metallic scatterer. The proposed approach not only estimates the position of the scatterer, but also determines the physical properties of an unknown metallic scatterer such as the shape (cylinder or sphere), the material (density), and the size (radius) in an anisotropic scattering case. Theoretical and numerical simulation results are also studied and investigated to show that the proposed approach can simultaneously characterize all those properties of an unknown metallic scatterer. Moreover, the advantage of the proposed approach is to avoid the complex iterative scheme in solving the direct scattering problem and results in smaller computational load and faster implementation.

The investigation of localized sound generation using two ultrasound beams.

A quantitative analysis of the effects of difference frequency, source separation, and crossing angle on the generated scattered difference frequency sound fields is presented to evaluate the feasibility of localized sound production using 2 uniform pistons. Nonlinear crossed beam experiments were also carried out in an anechoic chamber. Experimental results show that the audible sound could be generated within the interaction region defined by the overlap volume of 2 ultrasonic beams.