Critical Band Research Articles

Self-Weighted Semi-Supervised Classification for Joint EEG-Based Emotion Recognition and Affective Activation Patterns Mining

In Electroencephalography (EEG)-based affective brain-computer interfaces (aBCIs), there is a consensus that EEG features extracted from different frequency bands and channels have different abilities in emotion expression. Besides, EEG is so weak and non-stationary that easily causes distribution discrepancies for EEG data collected at different times; therefore, it is necessary to explore the affective activation patterns in cross-session emotion recognition. To address these two problems, we propose a self-weighted semi-supervised classification (SWSC) model in this paper for joint EEG-based cross-session emotion recognition and affective activation patterns mining, whose merits include 1) using both the labeled and unlabeled samples from different sessions for better capturing data characteristics, 2) introducing a self-weighted variable to learn the importance of EEG features adaptively and quantitatively, and 3) mining the activation patterns including the critical EEG frequency bands and channels automatically based on the learned self-weighted variable. Extensive experiments are conducted on the benchmark SEED_IV emotional data set and SWSC obtained excellent average accuracies of 77.40%, 79.55% and 81.52% in three cross-session emotion recognition tasks. Moreover, SWSC identifies that the Gamma frequency band contributes the most and the EEG channels in prefrontal, left/right temporal and (central) parietal lobes are more important for cross-session emotion recognition.

Tuning the mechanical properties of nanoglass-metallic glass composites with brick and mortar designs

We use molecular dynamics simulations to characterize the tensile deformation and failure of metallic glass (MG)-nanoglass (NG) composites designed with a brick and mortar architecture consisting of a 3 nm-grain-size NG matrix and a MG second phase. Results show that the overall strength of the matrix is significantly improved by the second phase without sacrificing the large NG matrix ductility. The mechanical synergy of the two phases is optimized by arranging the MG bricks in a staggered way, which effectively delocalizes the plastic deformation, hindering the buildup of local stress hot spots and the generation of critical shear bands.

Frequency Masking Effects for Vertical Whole-Body Vibration for Seated Subjects

Masking occurs when the perception of a stimulus is affected or covered by the presence of another signal in close proximity either in time or frequency. This study investigated frequency masking effects across a wide frequency range for whole-body vibration (WBV). The hypothesis that masking effects for WBV might be caused by sub-channels within the Pacinian channel was explored in two experiments. One experiment explored the masking effects of narrow band noise (NBN) on the perception threshold of sinusoidal vibrations; another explored the effect of different widths of NBN on the shift of the perception threshold for vertical vibration of seated subjects. The results show distinct masking effects for WBV based on frequency, albeit they do not support the existence of sub-channels within the Pacinian channel. Neither the typical masking effects associated with critical bands nor threshold shifts dependent on the bandwidth of the narrow band noise can be shown. Thus, the hypothesis does not appear to hold for WBV, but frequency masking must be considered for future studies and tactile applications.

Multi-objective long-short term memory recurrent neural networks for speech enhancement

Speech-in-noise perception is an important research problem in many real-world multimedia applications. The noise-reduction methods contributed significantly; however rely on a priori information about the noise signals. Deep learning approaches are developed for enhancing the speech signals in nonstationary noisy backgrounds and their benefits are evaluated for the perceived speech quality and intelligibility. In this paper, a multi-objective speech enhancement based on the Long-Short Term Memory (LSTM) recurrent neural network (RNN) is proposed to simultaneously estimate the magnitude and phase spectra of clean speech. During training, the noisy phase spectrum is incorporated as a target and the unstructured phase spectrum is transformed to its derivative that has an identical structure to corresponding magnitude spectrum. Critical Band Importance Functions (CBIFs) are used in training process to further improve the network performance. The results verified that the proposed multi-objective LSTM (MO-LSTM) successfully outscored the standard magnitude-aware LSTM (MA-LSTM), magnitude-aware DNN (MA-DNN), phase-aware DNN (PA-DNN), magnitude-aware GNN (MA-GNN) and magnitude-aware CNN (MA-CNN). Moreover, the proposed speech enhancement considerably improved the speech quality, intelligibility, noise-reduction and automatic speech recognition in changing noisy backgrounds, which is confirmed by the ANalysis Of VAriance (ANOVA) statistical analysis.

A Tandem Axial-Piston Unit Based Strategy for the Reduction of Noise Sources in Hydraulic Systems

This article presents a novel passive fluid borne noise source reduction strategy, based on tandem axial-piston unit indexing with the usage of symmetric lines. The strategy consists of setting the phase between the two synchronous units to accomplish destructive interference in targeted unit harmonics. A strategy capable of achieving destructive interference in all odd harmonics is investigated first analytically and then confirmed by a simulation study. Experiments on the proposed strategy confirmed its effectiveness at the first and third pump fundamental harmonics, and pressure ripple reduction was accomplished. The fluid borne noise source reduction in the first and third harmonic is verified to be propagated to pipe vibration and sound power. Regarding the first harmonic, pressure ripple was reduced by up to 18 dB; while for third harmonic, pressure ripple was reduced by up to 11 dB. In the experiment, however, noise cancellation is not achieved for the higher odd harmonics, as is instead found in the simulation. Conversely, transfer functions form pressure ripple to pipe wall acceleration are obtained experimentally, and a critical vibration band from 2000 Hz to 3000 Hz is identified as being crucial for effective overall sound power reduction.

Intrinsic and extrinsic effects on the brittle-to-ductile transition in metallic glasses

The effects of cooling rate, temperature, and applied strain rate on the tensile deformation behavior of a Cu64Zr36 metallic glass (MG) are investigated using large-scale molecular dynamics simulations. An increase in the quenching rate during sample preparation, as well as an increase of the temperature or the applied strain rate, affects the activation of shear transformation zones (STZs) and, consequently, the shear-banding processes, which ultimately causes a brittle-to-ductile transition in the deformation behavior of MGs. A quantitative interpretation for the observed enhanced ductility in MGs with an increasing quenching rate is obtained by sampling the saddle points on the potential energy surface. High quenching rates lead to lower energy barriers for activation of a local atomic rearrangement (STZ) as compared to those MGs obtained at low quenching rates. Although the glassy structure does not show significant variations with increasing temperature, the kinetic energy of the atoms increases dramatically, which allows the atoms to rearrange easily; therefore, the probability of homogeneous thermal activation of STZs increases. Finally, a large number of STZs can also be activated by deformation at high strain rates when a large amount of elastic energy is stored in the glassy matrix. Consequently, a high density of STZ events and, therefore, a more complex percolation process results in a low probability for strain localization and formation of critical shear bands. Our results provide an atomistic understanding for the strain localization mechanisms in metallic glasses and shed more light on the brittle-to-ductile transition.

Dynamic fracture behavior of Zr63Cu12Ni12Al10Nb3 metallic glass under high strain-rate loading

We investigate the fracture behavior of a Zr-based bulk metallic glass, Zr63Cu12Ni12Al10Nb3, subjected to quasi-static compression, dynamic compression and flyer-plate impact experiments. The metallic glasses yield once the stress reaches the critical value and subsequently premature fracture occurs in a brittle manner through shear bands expanding rapidly under uniaxial stress loading, leaving vein patterns, some of which evolves into dendrite-like patterns due to the increase of strain rate. For uniaxial strain loading, the equivalent stress is lower than critical stress and shear bands are suppressed. The fracture mechanisms are nucleation and coalescence of voids rather than shear bands evolution, exhibiting the typical fractography of ductile fracture with microvoids, dimples and cup-cone structures. Compared with quasi-static and dynamic compression, higher impact stress induced the reduction of available free volume under planar impact loading. In such a case, decohesion strength, as a key parameter which characters ductile fracture, becomes higher, resulting in change from shear banding to void coalescence under planar impact loading.

Some results on higher eigenvalue optimization

In this paper we obtain several results concerning the optimization of higher Steklov eigenvalues both in two and higher dimensional cases. We first show that the normalized (by boundary length) k-th Steklov eigenvalue on the disk is not maximized for a smooth metric on the disk for $$k\ge 3$$ . For $$k=1$$ the classical result of Weinstock (J Ration Mech Anal 3:745–753, 1954) shows that $$\sigma _1$$ is maximized by the standard metric on the round disk. For $$k=2$$ it was shown by Girouard and Polterovich (Funct Anal Appl 44(2):106–117, 2010) that $$\sigma _2$$ is not maximized for a smooth metric. We also prove a local rigidity result for the critical catenoid and the critical Mobius band as free boundary minimal surfaces in a ball under $$C^2$$ deformations. We next show that the first k Steklov eigenvalues are continuous under certain degenerations of Riemannian manifolds in any dimension. Finally we show that for $$k\ge 2$$ the supremum of the k-th Steklov eigenvalue on the annulus over all metrics is strictly larger that that over $$S^1$$ -invariant metrics. We prove this same result for metrics on the Mobius band.

Analysis of the crack development and shear transfer mechanisms of reinforced concrete beams with low amounts of shear reinforcement

Although a sufficient quantity of shear reinforcement can improve the shear resistance of reinforced concrete beams, the shear failure of beams with a low amount of shear reinforcement is normally characterized by the formation of a single shear crack. In this case, it is not possible to make the basic assumptions inherent in the parallel chord truss analysis or the compression-field approaches, namely, that distributed shears cracks form a pattern of quasi-parallel shear cracks. As an extension of an approach to describing the critical shear crack formation within the critical shear band, this paper presents an analysis of crack development and shear transfer mechanisms in beams with a low amount of shear reinforcement. The starting point for the analysis was the shear cracking state, which was described in the previous stage of the development for beams without shear reinforcement. The shear crack development and shear resisting mechanisms were analyzed using a group of appropriate equilibrium conditions, geometric conditions and constitutive relationships. The proposed method was validated by comparing the predicted shear resistance with the test results for rectangular, simply supported beams subjected to point loads included in the ACI-DAfStb databases. Furthermore, the validated analysis was used to investigate the relative contributions of different shear resisting mechanisms during the shear crack development. Through regression analyses of the concrete contribution and the ultimate crack length, a simplified model for the shear resistance of rectangular beams with a low mount of shear reinforcement was obtained. Predictions using both the analytical and simplified models are in good agreement with the test results, providing suitable tools for the transition from beams without reinforcement to those with high quantities of shear reinforcement.

Tuning the Mechanical Properties of Shape Memory Metallic Glass Composites with Brick and Mortar Designs

We use molecular dynamics simulations to characterize the tensile deformation and failure of shape memory alloy-bulk metallic glass composites (BMGCs). As the failure in BMGCs shifts from the MG matrix to the second phase different strategies are required to further improve their mechanical properties. Results show that a staggered brick BMGC displays enhanced ductility with mild strength loss following a homogeneously distributed plasticity. This demonstrates that judiciously designed BMGCs can effectively synergize the mechanical properties of metallic glasses and shape memory alloys by effectively distributing stress and preventing the generation of critical shear bands and premature failure while preserving strength.

Abnormal Acoustic Event Detection Based on Orthogonal Matching Pursuit in Security Surveillance System

This paper presents a feature extraction approach for surveillance system aimed at achieving the automatic detection and recognition of public security events. The proposed approach first generates a Gabor dictionary based on the human auditory critical frequency bands, and then uses the orthogonal matching pursuit (OMP) algorithm to sparse abnormal audio signal. We select the optimal several important atoms from the Gabor dictionary and extract the scale, frequency, and translation parameters of the atoms to form the OMP feature. The performance of OMP feature is compared with traditional acoustic features and their joint features, using support vector machine (SVM) and random forest (RF) classifiers. Experiments have been performed to evaluate the effectiveness of the OMP feature for supplementing traditional acoustic features. The results show the superior performance classifier for abnormal acoustic event detection (AAED) is RF. Furthermore, the introduction of the combined features addresses the problems of low recognition accuracy and poor robustness for the surveillance system in practical applications.

Voice Pathology Detection based on Analysis of Modulation Spectrum in Critical Bands

The paper presents an approach to the analysis of the modulation spectrum of a voice signal, in which the primary acoustic analysis is performed in bands of unequal width. Nonuniform analysis corresponds to the psychoacoustic laws of human perception of sound information. In the context of the analysis of the modulation spectrum, the considered approach can significantly reduce the resulting number of parameters, which greatly simplifies the task of detecting pathological changes in the voice signal based on the analysis of the parameters of the modulation spectrum. For frequency decomposition of a signal into bands of unequal width, two methods are considered: 1) DFT with channel combination and 2) the use of an nonuniform filter bank. The first method is characterized by a fixed time window for the analysis of all frequency components, while in the second method the time-frequency analysis plan is consistent with the critical frequency scale of the barks. For each method, a practical signal analysis circuit has been developed and described. The paper presents the experimental data on the application of the developed schemes for the analysis of the modulation spectrum to the problem of detecting pathology in a speech signal. The parameters of the modulation spectrum acted as information signs for a classifier built on the basis of linear discriminant analysis. Three different voice bases were used in the experiment (in two cases, the pathology was neurological ALS disease (amyotrophic lateral sclerosis), and in the third case, diseases of the larynx). The parameters of the modulation spectrum obtained in the DFT-based scheme with channel combining turned out to be more preferable for classification with a small number of features, however, greater accuracy (with an increase in the number of features) made it possible to obtain the parameters obtainedin the scheme based on an unequal filter bank. In all cases, the obtained classifiers were highly accurate (more than 97%). The obtained results show that the use of nonuniform time-frequency representation is preferable in the case when the analyzed signal is a sustained vowel phonation, since it provides higher accuracy of pathology detection using fewer modulation parameters

Sound quality prediction for power coupling mechanism of HEV based on CEEMD-HT and RVM

The sound quality evaluation model based on complementary ensemble empirical mode decomposition (CEEMD), Hilbert transform (HT) and relevance vector machine (RVM) is proposed to predict the sound quality of acoustical signals at power coupling mechanism of a hybrid electric vehicle (HEV). The technique is applied with wave filtering and CEEMD of the acoustical signals acquired at power coupling mechanism of HEV to achieve the intrinsic mode function (IMF) components. Built upon this is the calculation of the instantaneous frequencies of the IMF components by HT. Then, the critical frequency bands are used as the weight to calculate the weighted energy values of the IMF components. The weighted energy values are used as the new inputs of the sound quality evaluation model. Afterward, the subjective evaluation experiment of the acoustical signals at the power coupling mechanism is carried out based on pairwise comparison method. The subjective evaluation values are used as the outputs of the evaluation model. Finally, the new evaluation model is established based on RVM. In addition, the second RVM model is built for sound quality evaluation with the psychoacoustic objective parameters as the inputs. In this paper, the sound samples acquired under steady- and unsteady-state operating conditions are tested, respectively, in two models to obtain the prediction results. The prediction result suggests that the prediction accuracy of the evaluation model based on CEEMD-HT is higher than the evaluation model based on psychoacoustic objective parameters, and the prediction accuracy of the steady sound samples is higher than the unsteady sound samples.

Probabilistic assessment of steel buildings installed with passive control devices under multi-hazard scenario of earthquake and wind

The multi-hazard effects on multi-story steel buildings equipped with energy dissipative passive vibration response control devices are investigated for their performance under earthquake- and wind-induced forces. The passive response control devices include steel bracings, viscous, and viscoelastic dampers. The buildings without and with the passive control devices are modeled as multi-degree of freedom (M-DOF) systems, with the seismic masses lumped at each floor level. The governing differential equations of motion for the uncontrolled and controlled buildings are solved by using Newmark’s time integration approach. The dynamic response quantities are compared in terms of statistical distribution, which include the statistics of top floor acceleration, inter-storey drift, column base shear, and top floor displacement, for the uncontrolled and controlled buildings subjected to earthquake- and wind-induced forces. Conditional probabilities of failure are determined by overlapping the probability density function curves and conducting fragility analysis through cumulative distribution function curves to assess the effectiveness of the control devices against the natural hazards. It is observed that the design of the passive control devices employing additional damping may account for the performance of the structures under the multi-hazard scenario. Moreover, the variability in either of the demands affects the failure probability of the passively controlled structures extensively. Therefore, the probability of failure for a structure experiencing high-amplitude seismic hazard and a maximum wind speed within the critical frequency band during the design life varies largely for different damper schemes, which calls upon careful selection for design of a structure considering such multi-hazard scenario.

EEG-Based Emotion Recognition Using Logistic Regression with Gaussian Kernel and Laplacian Prior and Investigation of Critical Frequency Bands

Emotion plays a nuclear part in human attention, decision-making, and communication. Electroencephalogram (EEG)-based emotion recognition has developed a lot due to the application of Brain-Computer Interface (BCI) and its effectiveness compared to body expressions and other physiological signals. Despite significant progress in affective computing, emotion recognition is still an unexplored problem. This paper introduced Logistic Regression (LR) with Gaussian kernel and Laplacian prior for EEG-based emotion recognition. The Gaussian kernel enhances the EEG data separability in the transformed space. The Laplacian prior promotes the sparsity of learned LR regressors to avoid over-specification. The LR regressors are optimized using the logistic regression via variable splitting and augmented Lagrangian (LORSAL) algorithm. For simplicity, the introduced method is noted as LORSAL. Experiments were conducted on the dataset for emotion analysis using EEG, physiological and video signals (DEAP). Various spectral features and features by combining electrodes (power spectral density (PSD), differential entropy (DE), differential asymmetry (DASM), rational asymmetry (RASM), and differential caudality (DCAU)) were extracted from different frequency bands (Delta, Theta, Alpha, Beta, Gamma, and Total) with EEG signals. The Naive Bayes (NB), support vector machine (SVM), linear LR with L1-regularization (LR_L1), linear LR with L2-regularization (LR_L2) were used for comparison in the binary emotion classification for valence and arousal. LORSAL obtained the best classification accuracies (77.17% and 77.03% for valence and arousal, respectively) on the DE features extracted from total frequency bands. This paper also investigates the critical frequency bands in emotion recognition. The experimental results showed the superiority of Gamma and Beta bands in classifying emotions. It was presented that DE was the most informative and DASM and DCAU had lower computational complexity with relatively ideal accuracies. An analysis of LORSAL and the recently deep learning (DL) methods is included in the discussion. Conclusions and future work are presented in the final section.

Tuning the Mechanical Properties of Shape Memory Metallic Glass Composites with Nature Inspired Designs

We use large-scale molecular dynamics simulations of tensile loading to characterize the deformation and failure of nature inspired shape memory alloys-bulk metallic glass composite (BMGCs) designs. Results show synergy of the two phases resulting in heterogeneous distribution of stress assisted delocalization of plastic deformation. A staggered brick BMGC model displays enhanced ductility with mild strength loss following a rather homogeneous distribution of plastic events. These results reveal that nature inspired BMGC designs can effectively synergize the mechanical properties of metallic glasses and shape memory alloys by delaying the generation of critical shear bands and premature failure while preserving strength.

Aspect ratio effects on the serrated flow dynamic of TiZrHfCuNiBe high entropy metallic glass

The effects of aspect ratio on serrated flow dynamic and plasticity of TiZrHfCuNiBe high entropy bulk metallic glass were investigated. With increasing of the aspect ratio, the stress drop time decreases and a ductile-to-brittle transition was observed, which indicates the motion of shear bands will berestricted in the specimens with smaller aspect ratio. Thus, the shear band velocity can be estimated based on the stick-slip dynamics, which enhances with increasing of the aspect ratio, while the critical shear band velocity is almost identical. Therefore, the specimens with larger aspect ratio will achieve the critical shear band velocity faster until fracture. When the aspect ratio is relatively small, the value of shear-band stability index is larger and promotes stable propagation of the shear bands. In addition, the plasticity and serration size are not monotonic correlated. Benefitting from the abundant secondary shear bands and the interactions at various scales, the plastic deformation can be regarded as the self-organized critical state for samples with better plasticity and lower aspect ratio. However, the self-organized critical state will switch to a chaotic state with increasing of the aspect ratio.

Resonant Tunneling Spectroscopy to Probe the Giant Stark Effect in Atomically Thin Materials.

Each atomic layer in van der Waals heterostructures possesses a distinct electronic band structure that can be manipulated for unique device operations. In the precise device architecture, the subtle but critical band splits by the giant Stark effect between atomic layers, varied by the momentum of electrons and external electric fields in device operation, has not yet been demonstrated or applied to design original devices with the full potential of atomically thin materials. Here, resonant tunneling spectroscopy based on the negligible quantum capacitance of 2D semiconductors in resonant tunneling transistors is reported. The bandgaps and sub-band structures of various channel materials could be demonstrated by the new conceptual spectroscopy at the device scale without debatable quasiparticle effects. Moreover, the band splits by the giant Stark effect in the channel materials could be probed, overcoming the limitations of conventional optical, photoemission, and tunneling spectroscopy. The resonant tunneling spectroscopy reveals essential and practical information for novel device applications.

Electronic structure of bulk and multilayer black phosphorus under strain: a minimal model study

The tight-binding method is an important theoretical tool to investigate the physics of black phosphorus. Recently, a fifteen-parameter tight-binding model was proposed to precisely describe the band features. The large number of parameters is an obstacle for understanding the physics, and a minimal model for black phosphorus that grasps the main physics with satisfied precision is quite valuable. We use a tight banding model with only three parameters, that can reproduce the correct energy gap for monolayer, few-layer and bulk black phosphorus, to study the band structure black phosphorus under strain. The energy gap can be decreased by tensile strain normal to the phosphorus layers or in-plane compressive strain. At a certain critical strain, the gap of the phosphorus (either bulk, multilayer, or monolayer) is closed, and the dispersion is linear in the armchair direction and is parabolic in the other two principle directions. The simplicity of our model allows us to analytically investigate the energy gap and the critical strain as functions of the layer number of multilayer black phosphorus. When the strain exceeds the critical value, the valence and conduction bands evolve into two tents with their ridges touching each other.

Twin identification from speech: linear and non-linear cepstral features and models

Identifying a speaker from the speech by using computer-based learning techniques is normally in vogue for many years in research perspective. Twin identification from visual features is a challenging work specifically to identify a twin from identical twin pairs because they look alike in exhibiting manners, habits and so on. Speech can be used as a biometric to perform twin identification by imbibing adequate training imparted considering the fact that their vocal tract structure would be different in producing speech sounds. This work mainly highlights the usage of linear and non-linear frequency based Cepstral features and modelling methods to estimate the performance of the twin identification system. Set of speech utterances are separated into two sets namely training and testing. Utterances considered for training are concatenated and by performing the conventional pre-processing and appropriate feature extraction techniques, the proposed features are extracted. The feature vectors are normalized and given to the modeling techniques to create templates. Proposed features extracted from the utterances considered for testing are applied to the models and based on the classification criteria, twin is identifiedand the accuracy is categorized as sub-optimal and true success rates. Perceptual features with filters spaced in Equivalent Rectangular Bandwidth (ERB) for critical band provides better accuracy for the individual classification and decision level fusion classification.