arXiv subjects

Md. Rabiul Islam

Publications and source records attributed to Md. Rabiul Islam.

13 recordsLinked to original sources

arXiv2024

Masked Autoencoder with Swin Transformer Network for Mitigating Electrode Shift in HD-EMG-based Gesture Recognition

Multi-channel surface Electromyography (sEMG), also referred to as high-density sEMG (HD-sEMG), plays a crucial role in improving gesture recognition performance for myoelectric control. Pattern recognition models developed based on HD-sEMG, however, are vulnerable to changing recording conditions (e.g., signal variability due to electrode shift). This has resulted in significant degradation in performance across subjects, and sessions. In this context, the paper proposes the Masked Autoencoder with Swin Transformer (MAST) framework, where training is performed on a masked subset of HDsEMG channels. A combination of four masking strategies, i.e., random block masking; temporal masking; sensor-wise random masking, and; multi-scale masking, is used to learn latent representations and increase robustness against electrode shift. The masked data is then passed through MAST's three-path encoder-decoder structure, leveraging a multi-path Swin-Unet architecture that simultaneously captures time-domain, frequency-domain, and magnitude-based features of the underlying HD-sEMG signal. These augmented inputs are then used in a self-supervised pre-training fashion to improve the model's generalization capabilities. Experimental results demonstrate the superior performance of the proposed MAST framework in comparison to its counterparts.

eess.SP↗

arXiv2024

Collapsing scenarios of K-essence generalized Vaidya spacetime under $f(\bar{R},\bar{T})$ gravity

The paper investigates the collapse of the generalized emergent Vaidya spacetime in the setting of $f(\bar{R}, \bar{T})$ gravity, specifically in K-essence theory. In this study, the Dirac-Born-Infeld type non-standard Lagrangian is used to calculate the emergent metric $\bar{G}_{μν}$, which is not conformally equivalent to the conventional gravitational metric. We use the function $f(\bar{R}, \bar{T})$ to reflect the additive nature of the emergent Ricci scalar ($\bar{R}$) and the trace of the emergent energy-momentum tensor ($\bar{T}$). Our study demonstrates that certain choices of $f(\bar{R}, \bar{T})$ may result in the existence of a naked singularity caused by gravitational collapse. The alternative $f(\bar{R}, \bar{T})$ values resulted in an accelerating universe dominated by dark energy. Moreover, the investigation showed the presence of both positive and negative masses, which might suggest the coexistence of dark matter and dark energy. Furthermore, for a given quantity of kinetic part of the K-essence scalar field, mass is completely changed into energy, meaning that spacetime is Minkowskian. The K-essence theory may also be employed as a dark energy framework and a basic gravitational theory, making it possible for researchers to investigate a wide ranges of cosmic phenomena.

gr-qc↗

arXiv2024

Surface EMG-Based Inter-Session/Inter-Subject Gesture Recognition by Leveraging Lightweight All-ConvNet and Transfer Learning

Gesture recognition using low-resolution instantaneous HD-sEMG images opens up new avenues for the development of more fluid and natural muscle-computer interfaces. However, the data variability between inter-session and inter-subject scenarios presents a great challenge. The existing approaches employed very large and complex deep ConvNet or 2SRNN-based domain adaptation methods to approximate the distribution shift caused by these inter-session and inter-subject data variability. Hence, these methods also require learning over millions of training parameters and a large pre-trained and target domain dataset in both the pre-training and adaptation stages. As a result, it makes high-end resource-bounded and computationally very expensive for deployment in real-time applications. To overcome this problem, we propose a lightweight All-ConvNet+TL model that leverages lightweight All-ConvNet and transfer learning (TL) for the enhancement of inter-session and inter-subject gesture recognition performance. The All-ConvNet+TL model consists solely of convolutional layers, a simple yet efficient framework for learning invariant and discriminative representations to address the distribution shifts caused by inter-session and inter-subject data variability. Experiments on four datasets demonstrate that our proposed methods outperform the most complex existing approaches by a large margin and achieve state-of-the-art results on inter-session and inter-subject scenarios and perform on par or competitively on intra-session gesture recognition. These performance gaps increase even more when a tiny amount (e.g., a single trial) of data is available on the target domain for adaptation. These outstanding experimental results provide evidence that the current state-of-the-art models may be overparameterized for sEMG-based inter-session and inter-subject gesture recognition tasks.

cs.CV↗

arXiv2022

$f(\bar{R}, L(X))$-gravity in the context of dark energy with power law expansion and energy conditions

The motto of this work is to generate a general formalism of $f(\bar{R}, L(X))-$gravity in the context of dark energy under the framework of the {\bf K-}essence emergent geometry with the Dirac-Born-Infeld (DBI) variety of action, where $\bar{R}$ is the familiar Ricci scalar, $L(X)$ is the DBI type non-canonical Lagrangian with $X={1\over 2}g^{μν}\nabla_μϕ\nabla_νϕ$ and $ϕ$ is the {\bf K-}essence scalar field. The emergent gravity metric $\bar{G}_{μν}$ and the well known gravitational metric $g_{μν}$ are not conformally equivalent. We have constructed a modified field equation using the metric formalism in $f(\bar{R}, L(X))$-gravity incorporating the corresponding Friedmann equations in the framework of the background gravitational metric which is of Friedmann-Lema{î}tre-Robertson-Walker (FLRW) type. The solution of modified Friedmann equations have been deduced for the specific choice of $f(\bar{R}, L(X))$, which is of Starobinsky-type, using power law expansion method. The consistency of the model with the accelerating phase of the Universe has been shown, when we restrict ourselves to consider the value of the dark energy density, as $\dotϕ^{2}=\frac{8}{9}=0.888 <1$, which indicates that the present Universe is dark energy dominated. Graphical plots for the energy density ($ρ$), pressure ($p$) and equation of state parameter ($ω$) w.r.t. time ($t$) based on parametric values are interestingly consistent with the dark energy domination and hence accelerating features. We also put some light on the corresponding energy conditions and constraints of the $f(\bar{R}, L(X))$ theory with one basic example.

gr-qc↗

arXiv2022

Semantic Web in Healthcare: A Systematic Literature Review of Application, Research Gap, and Future Research Avenues

Today, healthcare has become one of the largest and most fast-paced industries due to the rapid development of digital healthcare technologies. The fundamental thing to enhance healthcare services is communicating and linking massive volumes of available healthcare data. However, the key challenge in reaching this ambitious goal is letting the information exchange across heterogeneous sources and methods as well as establishing efficient tools and techniques. Semantic Web (SW) technology can help to tackle these problems. They can enhance knowledge exchange, information management, data interoperability, and decision support in healthcare systems. They can also be utilized to create various e-healthcare systems that aid medical practitioners in making decisions and provide patients with crucial medical information and automated hospital services. This systematic literature review (SLR) on SW in healthcare systems aims to assess and critique previous findings while adhering to appropriate research procedures. We looked at 65 papers and came up with five themes: e-service, disease, information management, frontier technology, and regulatory conditions. In each thematic research area, we presented the contributions of previous literature. We emphasized the topic by responding to five specific research questions. We have finished the SLR study by identifying research gaps and establishing future research goals that will help to minimize the difficulty of adopting SW in healthcare systems and provide new approaches for SW-based medical systems progress.

cs.CY↗

arXiv2022

Collapsing scenario for the k-essence emergent generalised Vaidya spacetime in the context of massive gravity's rainbow

In this paper, we study the collapsing scenario for the k-essence emergent Vaidya spacetime in the context of massive gravity's rainbow. For this study, we consider that the background metric is Vaidya spacetime in massive gravity's rainbow. We show that the k-essence emergent gravity metric resembles closely to the new type of generalized Vaidya massive gravity metric with the rainbow deformations for null fluid collapse where we consider the k-essence scalar field as a function solely of the advanced or the retarded time. The k-essence emergent Vaidya massive gravity rainbow mass function is also different. This new type k-essence emergent Vaidya massive gravity rainbow metric has satisfied the required energy conditions. The existence of the locally naked central singularity, the strength and the strongness of the singularities for the rainbow deformations of the k-essence emergent Vaidya massive gravity metric are the interesting outcomes of the present work.

gr-qc↗

arXiv2022

Human Gait Analysis using Gait Energy Image

Gait recognition is one of the most recent emerging techniques of human biometric which can be used for security based purposes having unobtrusive learning method. In comparison with other bio-metrics gait analysis has some special security features. Most of the biometric technique uses sequential template based component analysis for recognition. Comparing with those methods, we proposed a developed technique for gait identification using the feature Gait Energy Image (GEI). GEI representation of gait contains all information of each image in one gait cycle and requires less storage and low processing speed. As only one image is enough to store the necessary information in GEI feature recognition process is very easier than any other feature for gait recognition. Gait recognition has some limitations in recognition process like viewing angle variation, walking speed, clothes, carrying load etc. Our proposed method in the paper compares the recognition performance with template based feature extraction which needs to process each frame in the cycle. We use GEI which gives relatively all information about all the frames in the cycle and results in better performance than other feature of gait analysis.

cs.CV↗

arXiv2021

Acute Lymphoblastic Leukemia Detection from Microscopic Images Using Weighted Ensemble of Convolutional Neural Networks

Acute Lymphoblastic Leukemia (ALL) is a blood cell cancer characterized by numerous immature lymphocytes. Even though automation in ALL prognosis is an essential aspect of cancer diagnosis, it is challenging due to the morphological correlation between malignant and normal cells. The traditional ALL classification strategy demands experienced pathologists to carefully read the cell images, which is arduous, time-consuming, and often suffers inter-observer variations. This article has automated the ALL detection task from microscopic cell images, employing deep Convolutional Neural Networks (CNNs). We explore the weighted ensemble of different deep CNNs to recommend a better ALL cell classifier. The weights for the ensemble candidate models are estimated from their corresponding metrics, such as accuracy, F1-score, AUC, and kappa values. Various data augmentations and pre-processing are incorporated for achieving a better generalization of the network. We utilize the publicly available C-NMC-2019 ALL dataset to conduct all the comprehensive experiments. Our proposed weighted ensemble model, using the kappa values of the ensemble candidates as their weights, has outputted a weighted F1-score of 88.6 %, a balanced accuracy of 86.2 %, and an AUC of 0.941 in the preliminary test set. The qualitative results displaying the gradient class activation maps confirm that the introduced model has a concentrated learned region. In contrast, the ensemble candidate models, such as Xception, VGG-16, DenseNet-121, MobileNet, and InceptionResNet-V2, separately produce coarse and scatter learned areas for most example cases. Since the proposed kappa value-based weighted ensemble yields a better result for the aimed task in this article, it can experiment in other domains of medical diagnostic applications.

eess.IV↗

arXiv2020

Self-supervised Deep Learning for Reading Activity Classification

Reading analysis can give important information about a user's confidence and habits and can be used to construct feedback to improve a user's reading behavior. A lack of labeled data inhibits the effective application of fully-supervised Deep Learning (DL) for automatic reading analysis. In this paper, we propose a self-supervised DL method for reading analysis and evaluate it on two classification tasks. We first evaluate the proposed self-supervised DL method on a four-class classification task on reading detection using electrooculography (EOG) glasses datasets, followed by an evaluation of a two-class classification task of confidence estimation on answers of multiple-choice questions (MCQs) using eye-tracking datasets. Fully-supervised DL and support vector machines (SVMs) are used to compare the performance of the proposed self-supervised DL method. The results show that the proposed self-supervised DL method is superior to the fully-supervised DL and SVM for both tasks, especially when training data is scarce. This result indicates that the proposed self-supervised DL method is the superior choice for reading analysis tasks. The results of this study are important for informing the design and implementation of automatic reading analysis platforms.

cs.HC↗

arXiv2019

S-ConvNet: A Shallow Convolutional Neural Network Architecture for Neuromuscular Activity Recognition Using Instantaneous High-Density Surface EMG Images

The concept of neuromuscular activity recognition using instantaneous high-density surface electromyography (HD-sEMG) images opens up new avenues for the development of more fluid and natural muscle-computer interfaces. However, the existing approaches employed a very large deep convolutional neural network (ConvNet) architecture and complex training schemes for HD-sEMG image recognition, which requires the network architecture to be pre-trained on a very large-scale labeled training dataset, as a result, it makes computationally very expensive. To overcome this problem, we propose S-ConvNet and All-ConvNet models, a simple yet efficient framework for learning instantaneous HD-sEMG images from scratch for neuromuscular activity recognition. Without using any pre-trained models, our proposed S-ConvNet and All-ConvNet demonstrate very competitive recognition accuracy to the more complex state of the art for neuromuscular activity recognition based on instantaneous HD-sEMG images, while using a ~ 12 x smaller dataset and reducing learning parameters to a large extent. The experimental results proved that the S-ConvNet and All-ConvNet are highly effective for learning discriminative features for instantaneous HD-sEMG image recognition especially in the data and high-end resource constrained scenarios.

eess.SP↗

arXiv2009

Improvement of Text Dependent Speaker Identification System Using Neuro-Genetic Hybrid Algorithm in Office Environmental Conditions

In this paper, an improved strategy for automated text dependent speaker identification system has been proposed in noisy environment. The identification process incorporates the Neuro- Genetic hybrid algorithm with cepstral based features. To remove the background noise from the source utterance, wiener filter has been used. Different speech pre-processing techniques such as start-end point detection algorithm, pre-emphasis filtering, frame blocking and windowing have been used to process the speech utterances. RCC, MFCC, MFCC, MFCC, LPC and LPCC have been used to extract the features. After feature extraction of the speech, Neuro-Genetic hybrid algorithm has been used in the learning and identification purposes. Features are extracted by using different techniques to optimize the performance of the identification. According to the VALID speech database, the highest speaker identification rate of 100.000 percent for studio environment and 82.33 percent for office environmental conditions have been achieved in the close set text dependent speaker identification system.

cs.SD↗

arXiv2009

Codebook Design Method for Noise Robust Speaker Identification based on Genetic Algorithm

In this paper, a novel method of designing a codebook for noise robust speaker identification purpose utilizing Genetic Algorithm has been proposed. Wiener filter has been used to remove the background noises from the source speech utterances. Speech features have been extracted using standard speech parameterization method such as LPC, LPCC, RCC, MFCC, (delta)MFCC and (delta)(delta) MFCC. For each of these techniques, the performance of the proposed system has been compared. In this codebook design method, Genetic Algorithm has the capability of getting global optimal result and hence improves the quality of the codebook. Comparing with the NOIZEOUS speech database, the experimental result shows that 79.62 percent accuracy has been achieved.

cs.SD↗

arXiv2007

Smooth maps of a foliated manifold in a symplectic manifold

The immersions of a smooth manifold $M$ in a symplectic manifold $(N,σ)$ inducing a given closed form $ω$ on $M$ satisfy the $C^0$-dense $h$-principle in the space of all continuous maps which pull back the deRham cohomology class of $σ$ onto that of $ω$. In this paper we prove a foliated version of this result due to Gromov.

math.DG↗