Search

Acoustic resolution photoacoustic micros- copy (AR-PAM) can achieve deeper imaging depth in biological tissue, with the sacrifice of imaging resolution compared with optical resolution photoacoustic microscopy (OR-PAM). Here we aim to enhance the AR-PAM image quality towards OR-PAM image, which specifically includes the enhancement of imaging resolution, restoration of micro-vasculatures, and reduction of artifacts. To address this issue, a network (MultiResU-Net) is first trained as generative model with simulated AR-OR image pairs, which are synthesized with physical transducer model. Moderate enhancement results can already be obtained when applying this model to in vivo AR imaging data. Nevertheless, the perceptual quality is unsatisfactory due to domain shift. Further, domain transfer learning technique under generative adversarial network (GAN) framework is proposed to drive the enhanced image's manifold towards that of real OR image. In this way, perceptually convincing AR to OR enhancement result is obtained, which can also be supported by quantitative analysis. Peak Signal to Noise Ratio (PSNR) and Structural Similarity Index (SSIM) values are significantly increased from 14.74 dB to 19.01 dB and from 0.1974 to 0.2937, respectively, validating the improvement of reconstruction correctness and overall perceptual quality. The proposed algorithm has also been validated across different imaging depths with experiments conducted in both shallow and deep tissue. The above AR to OR domain transfer learning with GAN (AODTL-GAN) framework has enabled the enhancement target with limited amount of matched in vivo AR-OR imaging data.

Situation awareness (SA) has received much attention in recent years because of its importance for operators of dynamic systems. Electroencephalography (EEG) can be used to measure mental states of operators related to SA. However, cross-subject EEG-based SA recognition is a critical challenge, as data distributions of different subjects vary significantly. Subject variability is considered as a domain shift problem. Several attempts have been made to find domain-invariant features among subjects, where subject-specific information is neglected. In this work, we propose a simple but efficient subject matching framework by finding a connection between a target (test) subject and source (training) subjects. Specifically, the framework includes two stages: (1) we train the model with multi-source domain alignment layers to collect source domain statistics. (2) During testing, a distance is computed to perform subject matching in the latent representation space. We use a reciprocal exponential function as a similarity measure to dynamically select similar source subjects. Experiment results show that our framework achieves a state-of-the-art accuracy 74.32% for the Taiwan driving dataset. National Research Foundation (NRF) This research is supported by the National Research Foundation, Singapore under its International Research Centres in Singapore Funding Initiative.

This work investigated the use of an ensemble deep random vector functional link (edRVFL) network for electroencephalogram (EEG)-based driver fatigue recognition. Against the low feature learning capability of the edRVFL network from raw EEG signals, two strategies were exploited in this work. Specifically, the first one was to exploit the advantages of the feature extractor module in CNNs, i.e., use CNN features as the input of the edRVFL network. The second one was to improve the feature learning capability of the edRVFL network. An enhanced edRFVL network named FGloWD-edRVFL was proposed, in which four enhancements were implemented, including random forest-based Feature selection, Global output layer, Weighting and entropy-based Dynamic ensemble. The proposed FGloWD-edRVFL network was evaluated on the challenging cross-subject driver fatigue recognition tasks. The results indicated that the proposed model could boost the recognition performance, significantly outperforming all strong baselines. The step-wise analysis further demonstrated the effectiveness of the proposed enhancements in the edRVFL network. 2023 The Author(s) Open Access funding provided by the Qatar National Library. Scopus

Data clustering is an important method used to discover naturally occurring structures in datasets. One of the most popular approaches is the grid-based concept of clustering algorithms. This kind of method is characterized by a fast processing time and it can also discover clusters of arbitrary shapes in datasets. These properties allow these methods to be used in many different applications. Researchers have created many versions of the clustering method using the grid-based approach. However, the key issue is the right choice of the number of grid cells. This paper proposes a novel grid-based algorithm which uses a method for an automatic determining of the number of grid cells. This method is based on the kdist function which computes the distance between each element of a dataset and its kth nearest neighbor. Experimental results have been obtained for several different datasets and they confirm a very good performance of the newly proposed method. Published version The paper is financed under the program of the Polish Minister of Science and Higher Education under the name ”Regional Initiative of Excellence” in the years 2019-2022; project number 020/RID/2018/19; the amount of financing PLN 12,000,000.00.

Successive layers in convolutional neural networks (CNN) extract different features from input images. Applications of CNNs to detect abnormalities in the 2D images or 3D volumes of body organs have recently become popular. However, computer-aided detection of diseases using deep CNN is challenging due to the absence of a large set of training medical images/scans and the relatively small and hard to detect abnormalities. In this paper, we propose a method for normalizing 3D volumetric scans using the intensity profile of the training samples. This aids the CNN by creating a higher contrast around the abnormal region of interest in the scan. We use the CQ500 head CT dataset to demonstrate the validity of our method for detecting different acute brain hemorrhages such as subarachnoid hemorrhage (SAH), intraparenchymal hemorrhage (IPH), subdural hematoma (SDH), and intraventricular hemorrhage (IVH). We compare the proposed method with a baseline, two variants of the 3D VGGNet architectures, Resnet, and show that the proposed method achieves significant improvement in classification performance. For binary classification, we achieved the best F1 score of 0.96 (normal vs SAH), 0.93 (normal vs IPH), 0.98 (normal vs SDH), and 0.99 (normal vs IVH), and for four-class classification, we obtained an average F1 score of 0.77. Finally, we show a limitation of the proposed method while detecting varied abnormalities. The proposed method has applications for abnormality detection for different organs. Nanyang Technological University This work was supported by the Lee Kong Chian School of Medicine and Data Science and AI Research (DSAIR) Centre of Nanyang Technological University Singapore (Project Number ADH-11/2017-DSAIR). The work of Parasuraman Padmanabhan and Balázs Gulyás was supported by the Cognitive NeuroImaging Centre (CONIC) at Nanyang Technological University Singapore.

In the paper we develop an algorithm based on the Parzen kernel estimate for detection of sudden changes in 3-dimensional shapes which happen along the edge curves. Such problems commonly arise in various areas of computer vision, e.g., in edge detection, bioinformatics and processing of satellite imagery. In many engineering problems abrupt change detection may help in fault protection e.g. the jump detection in functions describing the static and dynamic properties of the objects in mechanical systems. We developed an algorithm for detecting abrupt changes which is nonparametric in nature and utilizes Parzen regression estimates of multivariate functions and their derivatives. In tests we apply this method, particularly but not exclusively, to the functions of two variables.

The Elman neural network (ElmanNN) is well-known for its capability of processing dynamic information, which has led to successful applications in stock forecasting. In this paper, we introduce direct input-to-output connections (DIOCs) into the ElmanNN and show that the proposed Elman neural network with DIOCs (Elman-DIOCs) significantly out-performs the original ElmanNN without such DIOCs. Four different global stock indices, i.e., the Shanghai Stock Exchange (SSE) Composite Index, the Korea Stock Price Index (KOSPI), the Nikkei 225 Index (Nikkei225), and the Standard & Poor's 500 Index (SPX), are used to demonstrate the affecacy of the Elman-DIOCs in time-series prediction. We systematically evaluate 8 models, depending whether or not there are hidden layer biases, whether or not there are output layer biases, and whether or not there are DIOCs. The experimental results show that DIOCs lead to much better prediction accuracy, while requiring fewer than a half of the hidden neurons. Take the SPX index, for example - the root mean squared error (RMSE) and the mean absolute error (MAE) of the Elman-DIOCs are improved by 44.2% and 41.1%, respectively, compared to the ElmanNN, and 65.6% and 60.8%, respectively, compared to the multi-layer perceptron (MLP). We argue that (1) DIOCs can always help to improve accuracy, while reducing network complexity and computational burden, as long as the problem at hand (either regression or classification) has linear components, and (2) most real-world applications contain linear components. Therefore DIOCs will be almost always beneficial in any types of neural networks for classification or regression. We also point out that in rare cases where the problem at hand is entirely nonlinear, DIOCs should not be used. This study is funded by Joint Research Fund for Overseas Chinese Scholars and Scholars in Hong Kong and Macao (Grant No. 61828601), Natural Science Foundation of Shanxi Province (Grant No. 201801D121141), and Provincial Program on Key Research Projects of Shanxi (Social Development Area) (Grant No. 201903D321003).

The training set consists of many features that influence the classifier in different degrees. Choosing the most important features and rejecting those that do not carry relevant information is of great importance to the operating of the learned model. In the case of data streams, the importance of the features may additionally change over time. Such changes affect the performance of the classifier but can also be an important indicator of occurring concept-drift. In this work, we propose a new algorithm for data streams classification, called Random Forest with Features Importance (RFFI), which uses the measure of features importance as a drift detector. The RFFT algorithm implements solutions inspired by the Random Forest algorithm to the data stream scenarios. The proposed algorithm combines the ability of ensemble methods for handling slow changes in a data stream with a new method for detecting concept drift occurrence. The work contains an experimental analysis of the proposed algorithm, carried out on synthetic and real data.

For EEG-based drowsiness recognition, it is desirable to use subject-independent recognition since conducting calibration on each subject is time-consuming. In this paper, we propose a novel Convolutional Neural Network (CNN)-Long Short-Term Memory (LSTM) model for subject-independent drowsiness recognition from single-channel EEG signals. Different from existing deep learning models that are mostly treated as black-box classifiers, the proposed model can “explain” its decisions for each input sample by revealing which parts of the sample contain important features identified by the model for classification. This is achieved by a visualization technique by taking advantage of the hidden states output by the LSTM layer. Results show that the model achieves an average accuracy of 72.97% on 11 subjects for leave-one-out subject-independent drowsiness recognition on a public dataset, which is higher than the conventional baseline methods of 55.42%-69.27%, and state-of-the-art deep learning methods. Visualization results show that the model has discovered meaningful patterns of EEG signals related to different mental states across different subjects.

Glaucoma is a chronic eye disease, which causes gradual vision loss and eventually blindness. Accurate glaucoma screening at early stage is critical to mitigate its aggravation. Extracting high-quality features are critical in training of classification models. In this paper, we propose a deep ensemble network with attention mechanism that detects glaucoma using optic nerve head stereo images. The network consists of two main sub-components, a deep Convolutional Neural Network that obtains global information and an Attention-Guided Network that localizes optic disc while maintaining beneficial information from other image regions. Both images in a stereo pair are fed into these sub-components, the outputs are fused together to generate the final prediction result. Abundant image features from different views and regions are being extracted, providing compensation when one of the stereo images is of poor quality. The attention-based localization method is trained in a weakly-supervised manner and only image-level annotation is required, which avoids expensive segmentation labelling. Results from real patient images show that our approach increases recall (sensitivity) from the state-of-the-art 88.89% to 95.48%, while maintaining precision and performance stability. The marked reduction in false-negative rate can significantly enhance the chance of successful early diagnosis of glaucoma.

This paper describes an open access electroencephalography (EEG) data set for multitasking mental workload activity induced by a single-session simultaneous capacity (SIMKAP) experiment with 48 subjects. To validate the database, EEG spectral activity was evaluated with EEGLAB and the significant channels and activities for the experiment are highlighted. Classification performance was evaluated by training a support vector regression model on selected features from neighborhood component analysis based on a nine-point workload rating scale. With a reduced feature dimension, 69% classification accuracy was obtained for 3 identified workload levels from the rating scale with Cohen's kappa of 0.46. Accurate discrimination of mental workload is a desirable outcome in the field of operator performance analysis and BCI development; thus, we hope that our provided database and analyses can contribute to future investigations in this research field.

We study the three-dimensional turbulent Kolmogorov flow, i.e. the Navier-Stokes equations forced by a low-single-wave-number sinusoidal force in a periodic domain, by means of direct numerical simulations. This classical model system is a realization of anisotropic and non-homogeneous hydrodynamic turbulence. Boussinesq's eddy viscosity linear relation is checked and found to be approximately valid over half of the system volume. A more general nonlinear quadratic Reynolds stress development is proposed and its parameters estimated at varying the Taylor scale-based Reynolds number in the flow up to the value 200. The case of a forcing with a different shape, here chosen Gaussian, is considered and the differences with the sinusoidal forcing are emphasized., Comment: 11 pages

The statistical properties of species undergoing chemical reactions in a turbulent environment are studied. We focus on the case of reversible multi-component reactions of second and higher orders, in a condition close to chemical equilibrium sustained by random large-scale reactant sources, while the turbulent flow is highly developed. In such a state a competition exists between the chemical reaction that tends to dump reactant concentration fluctuations and enhance their correlation intensity and the turbulent mixing that on the contrary increases fluctuations and remove relative correlations. We show that a unique control parameter, the Damkh\"{o}ler number ($Da_\theta$) that can be constructed from the scalar Taylor micro-scale, the reactant diffusivity and the reaction rate characterises the functional dependence of fluctuations and correlations in a variety of conditions, i.e., at changing the reaction order, the Reynolds and the Schmidt numbers. The larger is such a Damkh\"{o}ler number the more depleted are the scalar fluctuations as compared to the fluctuations of a passive scalar field in the same conditions, and vice-versa the more intense are the correlations. A saturation in this behaviour is observed beyond $Da_\theta \simeq \mathcal{O}(10)$. We provide an analytical prediction for this phenomenon which is in excellent agreement with direct numerical simulation results., Comment: 17 pages , 9 figures

This paper proposes the deep convolution and correlated manifold embedded distribution alignment (DC-CMEDA) model, which is able to realize the transfer learning classification between and among various small datasets, and greatly shorten the training time. First, pre-trained Resnet50 network is used for feature transfer to extract smoke features because of the difficulty in training small dataset of forest fire smoke; second, a correlated manifold embedded distribution alignment (CMEDA) is proposed to register the smoke features in order to align the input feature distributions of the source and target domains; and finally, a trainable network model is constructed. This model is evaluated in the paper based on satellite remote sensing image and video image datasets. Compared with the deep convolutional integrated long short-term memory (DC-ILSTM) network, DC-CMEDA has increased the accuracy of video images by 1.50 %, and the accuracy of satellite remote sensing images by 4.00 %. Compared the CMEDA algorithm with the ILSTM algorithm, the number of iterations of the former has decreased to 10 times or less, and the algorithm complexity of CMEDA is lower than that of ILSTM. DC-CMEDA has a great advantage in terms of convergence speed. The experimental results show that DC-CMEDA can solve the problem of small sample smoke dataset detection and recognition.

An affective brain-computer interface (aBCI) is a direct communication pathway between human brain and computer, via which the computer tries to recognize the affective states of its user and respond accordingly. As aBCI introduces personal affective factors into human-computer interaction, it could potentially enrich the user’s experience during the interaction. Successful emotion recognition plays a key role in such a system. The state-of-the-art aBCIs leverage machine learning techniques which consist in acquiring affective electroencephalogram (EEG) signals from the user and calibrating the classifier to the affective patterns of the user. Many studies have reported satisfactory recognition accuracy using this paradigm. However, affective neural patterns are volatile over time even for the same subject. The recognition accuracy cannot be maintained if the usage of aBCI prolongs without recalibration. Existing studies have overlooked the performance evaluation of aBCI during long-term use. In this paper, we propose SAFE—an EEG dataset for stable affective feature selection. The dataset includes multiple recording sessions spanning across several days for each subject. Multiple sessions across different days were recorded so that the long-term recognition performance of aBCI can be evaluated. Based on this dataset, we demonstrate that the recognition accuracy of aBCIs deteriorates when re-calibration is ruled out during long-term usage. Then, we propose a stable feature selection method to choose the most stable affective features, for mitigating the accuracy deterioration to a lesser extent and maximizing the aBCI performance in the long run. We invite other researchers to test the performance of their aBCI algorithms on this dataset, and especially to evaluate the long-term performance of their methods.

The rapid advancements in machine learning, graphics processing technologies and the availability of medical imaging data have led to a rapid increase in the use of deep learning models in the medical domain. This was exacerbated by the rapid advancements in convolutional neural network (CNN) based architectures, which were adopted by the medical imaging community to assist clinicians in disease diagnosis. Since the grand success of AlexNet in 2012, CNNs have been increasingly used in medical image analysis to improve the efficiency of human clinicians. In recent years, three-dimensional (3D) CNNs have been employed for the analysis of medical images. In this paper, we trace the history of how the 3D CNN was developed from its machine learning roots, we provide a brief mathematical description of 3D CNN and provide the preprocessing steps required for medical images before feeding them to 3D CNNs. We review the significant research in the field of 3D medical imaging analysis using 3D CNNs (and its variants) in different medical areas such as classification, segmentation, detection and localization. We conclude by discussing the challenges associated with the use of 3D CNNs in the medical imaging domain (and the use of deep learning models in general) and possible future trends in the field., Comment: Published in Sensors Journal (https://www.mdpi.com/1424-8220/20/18/5097)

Extracting intra- and inter-subject parameters from Electroencephalogram (EEG) representing different Situation Awareness (SA) status is a critical challenge for objective SA recognition. Most of the existing work focuses on the subject-dependent classification that applies power spectrum density (PSD) features. In this paper, we propose a novel spectral-spatial (S-S) model for cross-subject fatigue-related SA recognition. The S-S model not only considers the biological topology across different brain regions to capture both local and global relations among different EEG channels, but also extracts spectral features for each EEG channel. Specifically, we firstly model the topological structure of EEG channels via an adjacency matrix which is built based on the Euclidean distance between EEG channels. Then, the graph convolution operation is employed to perform the neighbourhood aggregation for extracting spatial features. We test our model on a public dataset collected during driver's task performance. The subject-independent performance of the model is explored. Results demonstrate (1) the superior performance of our model compared with the state-of-the-art models on SA recognition from EEG signals. Specifically, our S-S model achieves 70.6% accuracy which is higher than traditional machine learning methods by 2.7%-6.8% and deep learning methods by 10.3%-11.6%; (2) EEG signal at the occipital region can better reflect the change of SA. National Research Foundation (NRF) Accepted version This research is supported by the National Research Foundation, Singapore under its International Research Centres in Singapore Funding Initiative. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not reflect the views of National Research Foundation, Singapore.

In turbulent combustion simulations, the flow structure at the unresolved scale level needs to be reasonably modeled. Following the idea of turbulent flamelet equation for the non-premixed flame case, which was derived based on the filtered governing equations (L. Wang, Combust. Flame 175 , 259 (2017)), the scalar dissipation term for tabulation can be directly computed from the resolved flowing quantities, instead of solving species transport equations. Therefore, the challenging source term closure for the scalar dissipation or any assumed probability density functions can be avoided; meanwhile the chemical sources are closed by scaling relations. The general principles are discussed in the context of large eddy simulation with case validation. The new model predictions of the bluff-body flame show sufficiently improved results, compared with these from the classic progress-variable approach.

The present work focuses on the flame–wall interaction (FWI) based on direct numerical simulations (DNS) of a head-on premixed flame quenching configuration at the statistically stationary state. The effects of FWI on the turbulent flame temperature, wall heat flux, flame dynamics and flow structures were investigated. In turbulent head-on quenching, particularly for high turbulence intensity, the distorted flames generally consist of the head-on flame part and the entrained flame part. The flame properties are jointly influenced by turbulence, heat generation from chemical reactions and heat loss to the cold wall boundary. For the present FWI configuration, as the wall is approached, the ‘influence zone’ can be identified as the region within which the flame temperature, scalar gradient and flame dilatation start to decrease, whereas the wall heat flux tends to increase. As the distance to the wall drops below the flame-quenching distance, approximately where the wall heat flux reaches its maximum value, chemical reactions become negligibly weak inside the ‘quenching zone’. A simplified counter-flow model is also proposed. With the reasonably proposed relation between the flame speed and the flame temperature, the model solutions match well with the DNS results, both qualitatively and quantitatively. Moreover, near-wall statistics of some important flame properties, including the flame dilatation, reaction progress variable gradient, tangential strain rate and curvature were analysed in detail under different wall boundary conditions.

Flame–wall interaction (FWI) in premixed turbulent combustion has been analysed based on a counter-flow like configuration at the statistically stationary state. For the present configuration, the ...

Direct numerical simulations (DNS) of oblique wall quenching of a turbulent V-flame and head-on quenching (HOQ) of a statistically planar flame by isothermal inert walls have been utilized to analy...

The tremendous success of machine learning algorithms at image recognition tasks in recent years intersects with a time of dramatically increased use of electronic medical records and diagnostic imaging. This review introduces the machine learning algorithms as applied to medical image analysis, focusing on convolutional neural networks, and emphasizing clinical aspects of the field. The advantage of machine learning in an era of medical big data is that significant hierarchal relationships within the data can be discovered algorithmically without laborious hand-crafting of features. We cover key research areas and applications of medical image classification, localization, detection, segmentation, and registration. We conclude by discussing research obstacles, emerging trends, and possible future directions. Published version

The local and non-local statistics of vorticity (i.e. curl of velocity vector ) and pseudo-vorticity (i.e. curl of density-weighted velocity vector ) have been investigated in detail based on a direct numerical simulation (DNS) database of freely propagating statistically planar flames in different regimes of premixed turbulent combustion. It has been found that the relative orientation of can be significantly different to that of for small values of the Karlovitz number and large values of the Reynolds number. The vorticity shows, depending on the case, considerable alignment with either the most extensive or compressive principal strain rate. In contrast, has been found to align with the intermediate principal strain rate for all cases. The transport statistics of Reynolds averaged enstrophy (i.e. ) and pseudo-enstrophy (i.e. ) have been investigated in detail. The non-local vorticity statistics in turbulent premixed combustion have been investigated based on the vorticity line segment (VLS) str...

Results related to the filtered non-premixed turbulent flame statistics are presented. Considering non-premixed turbulent combustion modeling in the framework of large eddy simulations (LES), a turbulent flamelet equation is derived based on the filtered governing equations. Using a direct numerical simulation (DNS) data set, the statistical properties of two important parameters, the scalar dissipation and chemical source term, are analyzed in details. Numerical results show that for the scalar dissipation term, the probability density function (PDF) of the magnitude of the mixture fraction gradient changes is strongly dependent on the filter scale, whereas the PDF of the turbulent diffusivity is almost invariant. Theoretically from the scalar structure function a scaling law can be expected for the conditional mean of the difference between the filtered and nonfiltered scalars with respect to the filter scale, which is important for estimating the chemical source in the subgrid scale models.

The necessity of soft computing techniques for medical image analysis is increasing every day. With new types of diseases affecting human beings, the diagnosis process becomes extremely important for efficient treatment planning. Soft computing techniques play a significant role in diagnosis and its allied techniques. Medical practitioners/researchers are always on the search for soft computing approaches with high performance measures. This is the scope for in-depth research in the area of these approaches in the context of medical applications. This Special Section in IEEE Access is an ideal platform for showcasing the state-of-the-art methods in these areas. Four high-quality articles have been accepted for this Special Section.

A turbulent premixed flame anchored by a solid wall can reach a statistically stationary state. Overall the configuration is counterflowlike. Flame-wall interactions under such conditions are key to understanding combustion in a confined space and in developing wall flame models. A numerical study is presented.

The self-similarity of a passive scalar in homogeneous isotropic decaying turbulence is investigated by the method of line segments (M. Gauding et al., Physics of Fluids 27.9 (2015): 095102). The analysis is based on a highly resolved direct numerical simulation of decaying turbulence. The method of line segments is used to perform a decomposition of the scalar field into smaller sub-units based on the extremal points of the scalar along a straight line. These sub-units (the so-called line segments) are parameterized by their length $\ell$ and the difference $\Delta\phi$ of the scalar field between the ending points. Line segments can be understood as thin local convective-diffusive structures in which diffusive processes are enhanced by compressive strain. From DNS, it is shown that the marginal distribution function of the length~$\ell$ assumes complete self-similarity when re-scaled by the mean length $\ell_m$. The joint statistics of $\Delta\phi$ and $\ell$, from which the local gradient $g=\Delta\phi/\ell$ can be defined, play an important role in understanding the turbulence mixing and flow structure. Large values of $g$ occur at a small but finite length scale. Statistics of $g$ are characterized by rare but strong deviations that exceed the standard deviation by more than one order of magnitude. It is shown that these events break complete self-similarity of line segments, which confirms the standard paradigm of turbulence that intense events (which are known as internal intermittency) are not self-similar.

In recent years, many deep learning methods, allowed for a significant improvement of systems based on artificial intelligence methods. Their effectiveness results from an ability to analyze large labeled datasets. The price for such high accuracy is the long training time, necessary to process such large amounts of data. On the other hand, along with the increase in the number of collected data, the field of data stream analysis was developed. It enables to process data immediately, with no need to store them. In this work, we decided to take advantage of the benefits of data streaming in order to accelerate the training of deep neural networks. The work includes an analysis of two approaches to network learning, presented on the background of traditional stochastic and batch-based methods.

Affective brain-computer interface (aBCI) introduces personal affective factors to human-computer interaction. The state-of-the-art aBCI tailors its classifier to each individual user to achieve accurate emotion classification. A subject-independent classifier that is trained on pooled data from multiple subjects generally leads to inferior accuracy, due to the fact that electroencephalography patterns vary from subject to subject. Transfer learning or domain adaptation techniques have been leveraged to tackle this problem. Existing studies have reported successful applications of domain adaptation techniques on SEED dataset. However, little is known about the effectiveness of the domain adaptation techniques on other affective datasets or in a cross-dataset application. In this paper, we focus on a comparative study on several state-of-the-art domain adaptation techniques on two datasets: 1) DEAP and 2) SEED. We demonstrate that domain adaptation techniques can improve the classification accuracy on both datasets, but not so effective on DEAP as on SEED. Then, we explore the efficacy of domain adaptation in a cross-dataset setting when the data are collected under different environments using different devices and experimental protocols. Here, we propose to apply domain adaptation to reduce the intersubject variance as well as technical discrepancies between datasets, and then train a subject-independent classifier on one dataset and test on the other. Experiment results show that using domain adaptation technique in a transductive adaptation setting can improve the accuracy significantly by 7.25%-13.40% compared to the baseline accuracy where no domain adaptation technique is used. National Research Foundation (NRF) Accepted version This work was supported by the National Research Foundation, Prime Minister’s Office, Singapore, through Its International Research Centers in Singapore Funding Initiative.

In this paper, we analyze the sea surface temperature obtained from the global drifter program. The experimental Fourier power spectrum shows a two-decade power-law behavior as $E_{\theta}(f)\propto f^{-7/3}$ in the frequency domain. Dimensional argument suggests a two-dimensional-like Lagrangian forward cascade, in which the enstrophy dissipation $\epsilon_{\Omega}$ is involved. Using the Hilbert-Huang transform and multi-level segment analysis, the measured high-order statistics and the corresponding singularity spectrum confirm the existence of the intermittency with a measured intermittency parameter $\mu_{\theta}\simeq0.10$, which is much weaker than the prediction by the conventional structure function method., Comment: 19 pages with 9 figures