Convolutional Neural Networks or Vision Transformers: Who Will Win the Race for Action Recognitions in Visual Data?

Understanding actions in videos remains a significant challenge in computer vision, which has been the subject of several pieces of research in the last decades. Convolutional neural networks (CNN) are a significant component of this topic and play a crucial role in the renown of Deep Learning. Inspired by the human vision system, CNN has been applied to visual data exploitation and has solved various challenges in various computer vision tasks and video/image analysis, including action recognition (AR). However, not long ago, along with the achievement of the transformer in natural language processing (NLP), it began to set new trends in vision tasks, which has created a discussion around whether the Vision Transformer models (ViT) will replace CNN in action recognition in video clips. This paper conducts this trending topic in detail, the study of CNN and Transformer for Action Recognition separately and a comparative study of the accuracy-complexity trade-off. Finally, based on the performance analysis's outcome, the question of whether CNN or Vision Transformers will win the race will be discussed.

FMC2 model based perception grading for dark insurgent network analysis.

The burgeoning role of social network analysis (SNA) in various fields raises complex challenges, particularly in the analysis of dark and dim networks involved in illicit activities. Existing models like the stochastic block model (SBM), exponential graph model (EGM), and latent space model (LSM) are limited in scope, often only suitable for one-mode networks. This article introduces a novel fuzzy multiple criteria multiple constraint model (FMC2) tailored for community detection in two-mode networks, which are particularly common in dark networks. The proposed method quantitatively determines the relationships between nodes based on a probabilistic measure and uses distance metrics to identify communities within the network. Moreover, the model establishes fuzzy boundaries to differentiate between the most and least influential nodes. We validate the efficacy of FMC2 using the Noordin Terrorist dataset and conduct extensive simulations to evaluate performance metrics. The results demonstrate that FMC2 not only effectively identifies communities but also ranks influential nodes within them, contributing to a nuanced understanding of complex networks. The method promises broad applicability and adaptability, particularly in intelligence and security domains where identifying influential actors within covert networks is critical.

Analysis of the vulnerability estimation and neighbor value prediction in autonomous systems.

The security within autonomous systems (AS)s is one of the important measures to keep network users safe and stable from the various type of Distributed Denial of Service (DDoS) attacks. Similar to the other existing attack types Internet control message protocol (ICMP) based attacks are remained open challenge on the Internet environment. In this study, we have proposed a method to estimate the vulnerability of 600 AS provider edge (PE) routers by sending ICMP packets and predicted AS neighbor values using least square regression (LSR) approach. The results of our study show that 265 AS PE routers are vulnerable due to ICMP flood attack from the 600 ASs which were analyzed. Additionally, we have predicted that about 60% of total AS neighbors will be reduced in the next 3 years. Our results indicate that some ASs still did not deploy the firewall system in the boundary of their networks. Similarly, we also observed that the majority of ASs which expected to have less neighbor values in the next 3 years is due to change their routing paths to find adjacent paths.

Characterization of English Braille Patterns Using Automated Tools and RICA Based Feature Extraction Methods.

Braille is used as a mode of communication all over the world. Technological advancements are transforming the way Braille is read and written. This study developed an English Braille pattern identification system using robust machine learning techniques using the English Braille Grade-1 dataset. English Braille Grade-1 dataset was collected using a touchscreen device from visually impaired students of the National Special Education School Muzaffarabad. For better visualization, the dataset was divided into two classes as class 1 (1-13) (a-m) and class 2 (14-26) (n-z) using 26 Braille English characters. A position-free braille text entry method was used to generate synthetic data. N = 2512 cases were included in the final dataset. Support Vector Machine (SVM), Decision Trees (DT) and K-Nearest Neighbor (KNN) with Reconstruction Independent Component Analysis (RICA) and PCA-based feature extraction methods were used for Braille to English character recognition. Compared to PCA, Random Forest (RF) algorithm and Sequential methods, better results were achieved using the RICA-based feature extraction method. The evaluation metrics used were the True Positive Rate (TPR), True Negative Rate (TNR), Positive Predictive Value (PPV), Negative Predictive Value (NPV), False Positive Rate (FPR), Total Accuracy, Area Under the Receiver Operating Curve (AUC) and F1-Score. A statistical test was also performed to justify the significance of the results.

A robust approach for industrial small-object detection using an improved faster regional convolutional neural network.

With the increasing pace in the industrial sector, the need for a smart environment is also increasing and the production of industrial products in terms of quality always matters. There is a strong burden on the industrial environment to continue to reduce impulsive downtime, concert deprivation, and safety risks, which needs an efficient solution to detect and improve potential obligations as soon as possible. The systems working in industrial environments for generating industrial products are very fast and generate products rapidly, sometimes leading to faulty products. Therefore, this problem needs to be solved efficiently. Considering this problem in terms of faulty small-object detection, this study proposed an improved faster regional convolutional neural network-based model to detect the faults in the product images. We introduced a novel data-augmentation method along with a bi-cubic interpolation-based feature amplification method. A center loss is also introduced in the loss function to decrease the inter-class similarity issue. The experimental results show that the proposed improved model achieved better classification accuracy for detecting our small faulty objects. The proposed model performs better than the state-of-the-art methods.

A new affinity matrix weighted k-nearest neighbors graph to improve spectral clustering accuracy.

Researchers have thought about clustering approaches that incorporate traditional clustering methods and deep learning techniques. These approaches normally boost the performance of clustering. Getting knowledge from large data-sets is quite an interesting task. In this case, we use some dimensionality reduction and clustering techniques. Spectral clustering is gaining popularity recently because of its performance. Lately, numerous techniques have been introduced to boost spectral clustering performance. One of the most significant part of these techniques is to construct a similarity graph. We introduced weighted k-nearest neighbors technique for the construction of similarity graph. Using this new metric for the construction of affinity matrix, we achieved good results as we tested it both on real and artificial data-sets.

Human-in-the-Loop Predictive Analytics Using Statistical Learning.

The human-in-the-loop cyber-physical system provides numerous solutions for the challenges faced by the doctors or medical practitioners. There is a linear trend of advancement and automation in the medical field for the early diagnosis of several diseases. One of the critical and challenging diseases in the medical field is coma. In the medical research field, currently, the prediction of these diseases is performed only using the data gathered from the devices only; however, the human's input is much essential to accurately understand their health condition to take appropriate decision on time. Therefore, we have proposed a healthcare framework involving the concept of artificial intelligence in the human-in- the-loop cyber-physical system. This model works via a response loop in which the human's intention is concluded by gathering biological signals and context data, and then, the decision is interpreted to a system action that is recognizable to the human in the physical environment, thereby completing the loop. In this paper, we have designed a model for early prognosis of coma using the electroencephalogram dataset. In the proposed approach, we have achieved the best results using a statistical learning algorithm called autoregressive integrated moving average in comparison to artificial neural networks and long short-term memory models. In order to measure the efficiency of our model, we have used the root mean squared error (RMSE), mean absolute error (MAE), and mean squared error (MSE) value to evaluate the linear models as it gives the difference between the measured value and true or correct value. We have achieved the least possible error value for our dataset. To conduct this experiment, we used the dataset available in the phsyionet opensource community.

Work Related and Personal Predictors of Covid 19 transmission

The paper provides new evidence from a survey of 2000 individuals in the US and UK related to predictors of Covid-19 transmission. Specifically, it investigates work and personal predictors of transmission experience reported by respondents using regression models to better understand possible transmission pathways and mechanisms in the community. Three themes emerge from the analysis. Firstly, transport roles and travelling practices are significant predictors of infection. Secondly, evidence from the US especially shows union membership, consultation over safety measures and the need to use public transport to get to work are also significant predictors. This is interpreted as evidence of the role of deprivation and of reactive workplace consultations. Thirdly and finally, there is some, often weaker, evidence that income, car-owership, use of a shared kitchen, university degree type, risk-aversion, extraversion and height are predictors of transmission. The comparative nature of the evidence indicates that the less uniformly stringent nature of the US lockdown provides more information about both structural and individual factors that predict transmission. The evidence about height is discussed in the context of the aerosol transmission debate. The paper concludes that both structural and individual factors must be taken into account when predicting transmission or designing effective public health measures and messages to prevent or contain transmission. JEL CodesI1 I12 I14 I18

An object localization optimization technique in medical images using plant growth simulation algorithm.

The analysis of leukocyte images has drawn interest from fields of both medicine and computer vision for quite some time where different techniques have been applied to automate the process of manual analysis and classification of such images. Manual analysis of blood samples to identify leukocytes is time-consuming and susceptible to error due to the different morphological features of the cells. In this article, the nature-inspired plant growth simulation algorithm has been applied to optimize the image processing technique of object localization of medical images of leukocytes. This paper presents a random bionic algorithm for the automated detection of white blood cells embedded in cluttered smear and stained images of blood samples that uses a fitness function that matches the resemblances of the generated candidate solution to an actual leukocyte. The set of candidate solutions evolves via successive iterations as the proposed algorithm proceeds, guaranteeing their fit with the actual leukocytes outlined in the edge map of the image. The higher precision and sensitivity of the proposed scheme from the existing methods is validated with the experimental results of blood cell images. The proposed method reduces the feasible sets of growth points in each iteration, thereby reducing the required run time of load flow, objective function evaluation, thus reaching the goal state in minimum time and within the desired constraints.

Real-time Medical Emergency Response System: Exploiting IoT and Big Data for Public Health.

Healthy people are important for any nation's development. Use of the Internet of Things (IoT)-based body area networks (BANs) is increasing for continuous monitoring and medical healthcare in order to perform real-time actions in case of emergencies. However, in the case of monitoring the health of all citizens or people in a country, the millions of sensors attached to human bodies generate massive volume of heterogeneous data, called "Big Data." Processing Big Data and performing real-time actions in critical situations is a challenging task. Therefore, in order to address such issues, we propose a Real-time Medical Emergency Response System that involves IoT-based medical sensors deployed on the human body. Moreover, the proposed system consists of the data analysis building, called "Intelligent Building," depicted by the proposed layered architecture and implementation model, and it is responsible for analysis and decision-making. The data collected from millions of body-attached sensors is forwarded to Intelligent Building for processing and for performing necessary actions using various units such as collection, Hadoop Processing (HPU), and analysis and decision. The feasibility and efficiency of the proposed system are evaluated by implementing the system on Hadoop using an UBUNTU 14.04 LTS coreTMi5 machine. Various medical sensory datasets and real-time network traffic are considered for evaluating the efficiency of the system. The results show that the proposed system has the capability of efficiently processing WBAN sensory data from millions of users in order to perform real-time responses in case of emergencies.

A partition-based active contour model incorporating local information for image segmentation.

Active contour models are always designed on the assumption that images are approximated by regions with piecewise-constant intensities. This assumption, however, cannot be satisfied when describing intensity inhomogeneous images which frequently occur in real world images and induced considerable difficulties in image segmentation. A milder assumption that the image is statistically homogeneous within different local regions may better suit real world images. By taking local image information into consideration, an enhanced active contour model is proposed to overcome difficulties caused by intensity inhomogeneity. In addition, according to curve evolution theory, only the region near contour boundaries is supposed to be evolved in each iteration. We try to detect the regions near contour boundaries adaptively for satisfying the requirement of curve evolution theory. In the proposed method, pixels within a selected region near contour boundaries have the opportunity to be updated in each iteration, which enables the contour to be evolved gradually. Experimental results on synthetic and real world images demonstrate the advantages of the proposed model when dealing with intensity inhomogeneity images.