Search

Internet of Things (IoT) has brought changes in many fields by joining physical space with the cyber space. The IoT devices are becoming increasingly complex. With the rapid deployment of cameras, tasks in IoT like visual information are more important, but IoT devices have limited computing resources, including power, computing ability, storage, etc. Some tasks that might be perfectly normal to perform on a computer would be rather challenging on an IoT device. Therefore, how to maintain acceptable performance while minimizing resources is becoming a more consequential part in IoT. In the paper, we aim to solve the problem of object detection and tracking in IoT while minimizing resources. The traditional algorithms need to use convolutional neural network (CNN) to identify different objects in each frame, and then determine the tracking target from identified objects, which typically requires a lot of computing resources. By incorporating spatial attention, and multi-domain network, we proposed a novel algorithm named as Spatial Attention Powered Multi-Domain Network (SA-MDNet). By adding spatial attention mechanism to the original MDNet model, and using multi class cross entropy loss, we are able to distinguish the background and the target in different video sequences effectively and efficiently. This novel algorithm achieves similar performance on the OTB 50/100/2013 datasets compared to several state-of-art models, while uses only a fraction of the memory compared to MDNet.

Digital electronics course is one of the very fundamental courses for the students of undergraduate programme of electrical and electronic engineering (EEE) and the other undergraduate engineering disciplines. Therefore, "digital electronics" shall be taught effectively, so that students can apply the knowledge learned to solve their real-life engineering problems. A teacher needs to adopt new teaching methodologies to attract current generation of students, and thus, to prepare them with practical knowledge and skills. Skills in the cognitive domain of Bloom's taxonomy revolve around knowledge, comprehension and critical thinking of a particular topic. This makes teaching and learning more effective and efficient. In this paper, the teaching method of 'digital electronics' course for the undergraduate EEE students in the cognitive domain has been described with an example. Class performance evaluation in two different cohorts shows that the students' results improve after using this approach.Keywords: Bloom's taxonomy, cognitive domain, digital electronics course, teaching methods.

IoT has started capturing the market all over the world. Nearly all household appliances are available and consumer can buy them. Lack of security leaves the customer potentially exposed to various risk factors. Risk might be ranging daily routine data theft to great financial risk, even in some cases there is life threating risk for specific or group of persons. On analyzing various devices, through that come to know that many are not devices enforced strong passwords are not devices applied, or brute-force attacks is not protected, physical replacement of faulty sensors or wiretapping the traffic. As per approximation, every two among ten applications of mobile which are being used for managing and handling IOT devices are vulnerable and are not using SSL for communication data encryption. IoT involve billions of sensors spreading all around the world. Many of them are not physically protected. Attacker could access those sensors/devices and manipulate them or replace them. Captured data through sensors is also not safe if transmitted without encrypting them. Further many devices don’t have such power to perform heavy encryption. In this research Proposed a PUF (Physically unclonable function) and light weight encryption technique-based solution. Physically unclonable function create a class for new embedded and its security purpose. PUF mainly concentrate on shift the security challenges and issues. It has a very simple architecture and can solve problem of energy-constrained IoT devices. Through this research work introduced PUF functionality with hardware sensors. Since PUF function cannot be cloned therefore sensors are safe and any modification or replacement can be easily detected. Further data received from sensors will be encrypted using light weigh encryption technique. In future this proposed methodology explore a how to store safely vast amount of data generated through billions of sensors.

Functional Encryption (FE) expands traditional public-key encryption in two different ways: it supports fine-grained access control and allows learning a function of the encrypted data. In this paper, we review all FE classes, describing their functionalities and main characteristics. In particular, we mention several schemes for each class, providing their security assumptions and comparing their properties. To our knowledge, this is the first survey that encompasses the entire FE family., To appear in Advances in Mathematics of Communications, https://www.aimsciences.org/article/doi/10.3934/amc.2021049

Zero-Shot Learning (ZSL) aims to recognize unseen classes that never appear during training. Recently, generative adversarial networks (GANs) have been introduced to convert ZSL into a supervised learning problem by synthesizing unseen visual features. However, since unseen classes are never experienced for the generator during training, the synthesized unseen visual features often become heavily biased towards seen classes, or sometimes there is even no meaningful class that can be assigned to them. This is known as the \textit{bias problem}. In this paper, we propose a novel method, namely Adaptive Bias-Aware GAN (ABA-GAN), to alleviate generating biased visual features. For this purpose, we build a semantic adversarial network to regularize the feature generator. Specifically, an adaptive adversarial loss is proposed to constrain the feature distributions, which avoids the generation of meaningless visual features. Meanwhile, a domain divider is presented to explicitly distinguish synthesized visual features between seen and unseen domains, such that the bias towards seen classes can be alleviated. Moreover, we propose a novel metric named bias score (BS) to explicitly quantify the degree of the strong bias. Extensive experiments on four widely used benchmark datasets demonstrate that our proposed method outperforms the state-of-the-art approaches under both ZSL and GZSL protocols.

Deep learning methods have achieved significant progress in the presence of correctly annotated datasets in instance segmentation. However, object classes in large-scale datasets are sometimes ambiguous, which easily causes confusion. Besides, limited experience and knowledge of annotators can lead to mislabeled object semantic classes. To solve this issue, a novel method is proposed in this paper, which considers different roles of noisy class labels in different sub-tasks. Our method is based on two basic observations: firstly, the foreground-background annotation of a sample is correct even though its class label is noisy. Secondly, symmetric loss benefits the model robustness to noisy labels but harms the learning of hard samples, while cross entropy loss is the opposite. Based on the two basic observations, in the foreground-background sub-task, cross entropy loss is used to fully exploit correct gradient guidance. In the foreground-instance sub-task, symmetric loss is used to prevent incorrect gradient guidance provided by noisy class labels. Furthermore, we apply contrastive self-supervised loss to update features of all foreground, to compensate for insufficient guidance provided by partially correct labels especially in the highly noisy setting. Extensive experiments conducted with three popular datasets (i.e., Pascal VOC, Cityscapes and COCO) have demonstrated the effectiveness of our method in a wide range of noisy class label scenarios.

This article explains problems of individualization of education and extracurricular activities’ programs’ development. Main problems that need to be solved at various stages of the organization of extracurricular activities, from its methodological study to organizational and technical planning are analyzed. An overview of the directions of extracurricular activities, the forms in which it can be organized as well as the specifics of its implementation at various levels of education is offered. In order to reduce the labor intensity of time costs taking into account the time allocated for its implementation as well as the ability to generate plans taking into account individual preferences an intelligent information system was developed. Multi-criteria mathematical models from the class of optimization methods for solving the posed problem of decision support, functional and software-architectural developed model of a web service and examples of its work on the stages of planning general activities in the optimization formulation are presented. The proposed approach can be widely used in educational institutions of various levels interested in reducing the cost of organizing extracurricular work, automatic checking for compliance with requirements, as well as generating a work plan taking into account the opinions of various categories of experts.

Entity resolution (ER) identifies data instances referring to the same real-world entity and has received enormous research attention. In this paper, we examine the task of ER from a broader perspective, with its input extended from textual records, which are conventionally studied in the literature, to other modalities such as check-in sequences, GPS trajectories and surveillance video frames to generate new applications. Our goal in this paper is to design an effective model to uniformly support all these ER applications with different input formats. Technically, we fully exploit the semantic contexts of embedding vectors for the pair of input instances. In particular, we propose an integrated multi-context attention framework that takes into account self-attention, pair-attention and global-attention from three types of context. The idea can be further extended to incorporate attribute attention in order to support structured datasets. We conduct extensive experiments on a diverse class of entity resolutions tasks, including tasks on unstructured, structured and dirty textual records, check-in sequences, GPS trajectories and surveillance video frames. The experimental results verified the effectiveness and generality of our model. When compared with strong baselines in these applications, our model can achieve superior or comparative performance.

Class imbalance is a prevalent phenomenon in various real-world applications and it presents significant challenges to model learning, including deep learning. In this work, we embed ensemble learning into the deep convolutional neural networks (CNNs) to tackle the class-imbalanced learning problem. An ensemble of auxiliary classifiers branching out from various hidden layers of a CNN is trained together with the CNN in an end-to-end manner. To that end, we designed a new loss function that can rectify the bias toward the majority classes by forcing the CNN's hidden layers and its associated auxiliary classifiers to focus on the samples that have been misclassified by previous layers, thus enabling subsequent layers to develop diverse behavior and fix the errors of previous layers in a batch-wise manner. A unique feature of the new method is that the ensemble of auxiliary classifiers can work together with the main CNN to form a more powerful combined classifier, or can be removed after finished training the CNN and thus only acting the role of assisting class imbalance learning of the CNN to enhance the neural network's capability in dealing with class-imbalanced data. Comprehensive experiments are conducted on four benchmark data sets of increasing complexity (CIFAR-10, CIFAR-100, iNaturalist, and CelebA) and the results demonstrate significant performance improvements over the state-of-the-art deep imbalance learning methods.

Objectives: This study aimed to discover the extent to which medical students can evaluate medical websites, evaluation criteria used, factors affecting their abilities and whether a teaching intervention could rectify problems. Medical students and practitioners are required to evaluate medical information available on the Internet. Most current medical students are familiar with the Internet, but their ability to evaluate material may require improvement. Methods: A class of undergraduate medical students evaluated an unreliable medical website, received a teaching intervention on website evaluation criteria and re-evaluated the same site. This mixed-methods study was conducted at Sultan Qaboos University, Muscat, Oman, from September to December 2018. Results: A total of 149 (response rate: 82.3%) students participated. Students spent, on average, 4.69 hours per day on the Internet. No significant correlations were found between demographic indicators and Internet time. On a 10-point Likert scale, students’ scores ranged from 5–6, with no significant differences between the pre- and post-intervention evaluations, except for increased polarisation away from the mean. Qualitative comments indicated an awareness of relevant criteria but an overall inability to critically apply them. Conclusion: The results indicate that one cannot make a blanket statement about medical students’ ability to evaluate medical websites despite their familiarity with technology. Moreover, website evaluation should be viewed primarily from the information perspective and that critical thinking ability may play a major role. Due to these overriding factors, short interventions are unlikely to have an impact, and other educational strategies should be developed. These are necessary to ensure that medical students can function independently as life-long learners and medical professionals. Keywords: Internet; Medical Students; Oman.

Recently, a multitude of methods for image-to-image translation have demonstrated impressive results on problems, such as multidomain or multiattribute transfer. The vast majority of such works leverages the strengths of adversarial learning and deep convolutional autoencoders to achieve realistic results by well-capturing the target data distribution. Nevertheless, the most prominent representatives of this class of methods do not facilitate semantic structure in the latent space and usually rely on binary domain labels for test-time transfer. This leads to rigid models, unable to capture the variance of each domain label. In this light, we propose a novel adversarial learning method that: 1) facilitates the emergence of latent structure by semantically disentangling sources of variation and 2) encourages learning generalizable, continuous, and transferable latent codes that enable flexible attribute mixing. This is achieved by introducing a novel loss function that encourages representations to result in uniformly distributed class posteriors for disentangled attributes. In tandem with an algorithm for inducing generalizable properties, the resulting representations can be utilized for a variety of tasks such as intensity-preserving multiattribute image translation and synthesis, without requiring labeled test data. We demonstrate the merits of the proposed method by a set of qualitative and quantitative experiments on popular databases such as MultiPIE, RaFD, and BU-3DFE, where our method outperforms other state-of-the-art methods in tasks such as intensity-preserving multiattribute transfer and synthesis.

Zero-shot learning (ZSL) is a pretty intriguing topic in the computer vision community since it handles novel instances and unseen categories. In a typical ZSL setting, there is a main visual space and an auxiliary semantic space. Most existing ZSL methods handle the problem by learning either a visual-to-semantic mapping or a semantic-to-visual mapping. In other words, they investigate a unilateral connection from one end to the other. However, the connection between the visual space and the semantic space are bilateral in reality, that is, the visual space depicts the semantic space; the semantic space, on the other hand, describes the visual space. In this article, therefore, we investigate the bilateral connections in ZSL and present a novel model, called Boomerang-GAN, by taking advantage of conditional generative adversarial networks (GANs). Specifically, we generate unseen visual samples from their category semantic embeddings by a conditional GAN. Different from the existing generative ZSL methods that only consider generating visual features from class descriptions, our method also considers that the generated visual features can be translated back to their corresponding semantic embeddings by introducing a multimodal cycle-consistent loss. Extensive experiments of both ZSL and generalized ZSL on five widely used datasets verify that our method is able to outperform previous state-of-the-art approaches in both recognition and segmentation tasks.

Intelligent systems, like driver assistance systems, remain within vehicles and help drivers by providing essential information about traffic, blockage of roads, and possible routes for safe driving. The objective of scene text spotting in a driver assistance system is to localize and recognize scene texts, signs of milestones, traffic panels, and road marks in natural scene images. However, text edges get fainted due to adverse weather conditions, like fog, rain, smog, or poor contrast. This makes the task of spotting more challenging. In this paper, we propose an end-to-end trainable deep neural network, known as Met-MLTS, that can address the issue of spotting multi-oriented text instances in scene images captured in adverse meteorological conditions. It localizes words, predicts script class, and performs word spotting for every rotated bounding box. It is a fast multilingual scene text spotter that utilizes hierarchical spatial context, channel-wise inter-dependencies, and semantic edge supervision to localize and recognize words and predict script class in scene images using smartphones. We explore inter-class interference to reduce the misclassification problem. A light-weight recognition module for multilingual character segmentation, word-level recognition, and script identification is incorporated. We demonstrate the efficacy of our spotting network on resource-constraint devices.

We present an approach to analyse learning outcomes in a broad class of misspecified environments, spanning both single-agent and social learning. We introduce a novel “prediction accuracy” order over subjective models and observe that this makes it possible to partially restore standard martingale convergence arguments that apply under correctly specified learning. Based on this, we derive general conditions to determine when beliefs in a given environment converge to some long-run belief either locally or globally (i.e. from some or all initial beliefs). We show that these conditions can be applied, first, to unify and generalize various convergence results in previously studied settings. Second, they enable us to analyse environments where learning is “slow”, such as costly information acquisition and sequential social learning. In such environments, we illustrate that even if agents learn the truth when they are correctly specified, vanishingly small amounts of misspecification can generate extreme failures of learning.

API misuses are well-known causes of software crashes and security vulnerabilities. However, their detection and repair is challenging given that the correct usages of (third-party) APIs might be obscure to the developers of client programs. This paper presents the first empirical study to assess the ability of existing automated bug repair tools to repair API misuses, which is a class of bugs previously unexplored. Our study examines and compares 14 Java test-suite-based repair tools (11 proposed before 2018, and three afterwards) on a manually curated benchmark (APIREPBENCH) consisting of 101 API misuses. We develop an extensible execution framework (APIARTY) to automatically execute multiple repair tools. Our results show that the repair tools are able to generate patches for 28% of the API misuses considered. While the 11 less recent tools are generally fast (the median execution time of the repair attempts is 3.87 minutes and the mean execution time is 30.79 minutes), the three most recent are less efficient (i.e., 98% slower) than their predecessors. The tools generate patches for API misuses that mostly belong to the categories of missing null check, missing value, missing exception, and missing call. Most of the patches generated by all tools are plausible (65%), but only few of these patches are semantically correct to human patches (25%). Our findings suggest that the design of future repair tools should support the localisation of complex bugs, including different categories of API misuses, handling of timeout issues, and ability to configure large software projects. Both APIREPBENCH and APIARTY have been made publicly available for other researchers to evaluate the capabilities of repair tools on detecting and fixing API misuses.

The wireless sensor networks (WSNs) are special network which has purpose of gathering information in certain area. Multipath routing is the paramount path of addressing QoS and energy balancing concerns in low power and lossy networks (LLNs) especially in IoT technology. Most of the extant efforts bring the limited number of disjoint paths into play, and intersect the traffic among them pursuant to a compound metric or centralized optimization problem. This paper proposes a multi-class multipath routing protocol for LLNs (called M2RPL), that construct a braided multipath routing graph based on the standard RPL (IPv6 Routing Protocol for Low-Power and Lossy Networks) protocol, and an energy balanced optimal rate assignment mechanism (EBORA) that solves a local multi class optimization problem for minimum cost traffic rate assignment. Moreover the algorithm tries to maximize network lifetime by equalizing the energy dissipation rates of equi-level nodes. The simulation results expose the efficiency of the proposed framework, leading to an average 15% increase in lifetime, and improved QoS performance in terms of delay and reliability, compared to some well-known algorithms.

Exploring the characterization laws of image data and improving the efficiency of image data characterization knowledge is essential to promote the development of the Internet of Things technology. Considering that images in the real world usually contain multiple objects, and the objects are closely dependent. For these reasons, it brings great challenges to the robust representation learning of multi-label images. In general, researchers model the relationship between objects based on a class activation map and use graph convolution to mine the dependencies between objects. However, graph structure data often contain noise, which means that the edges between nodes are sometimes not so reliable, and the relative importance of neighbors is also different. Based on this, our goal is to reduce noisy connections and false connections between objects, eliminate multi-label image representation bias, and learn robust representations. Therefore, we propose a Robust Representation Learning method for multi-label images driven by Graph Attention Network (RRL-GAT). Specifically, to reduce the accidental false connection of objects in the image, we propose the Class Attention Graph convolution module (C-GAT) to mine the strong association structure between categories. Besides, for the dynamic correlation between objects in the image, we propose an Adaptive Graph Attention Convolution module (A-GAT) to capture the subtle dynamic dependencies in the image. The results on two authoritative datasets show that our method is significantly better than all current state-of-the-art methods. Besides, the visualization results show that RRL-GAT can capture the semantic relationship of a specific input image and has sufficient recognizability.

Internet of Things (IoT) and 5G network are fundamental building blocks for Industrial Internet of Things (IIoT). IoT has enabled real-time monitoring and actuation in industrial floors and machinery,aimed at improving the efficiency and safety of industrial activities and processes. On the other hand,5G networks will provide ultra-reliable and low-latency communication for the wireless integration of autonomous industrial machinery,mobile vehicles,and robots,and management systems,aimed at the real-time control and management of industrial machinery towards smart factories. In IIoT,machine learning (ML) will also play a fundamental role in handling complex tasks at industrial machinery and 5G networks management,configuration,and control. However,ML suffers from the cold-start problem and needs a large amount of highly accurate data samples for model training,which is costly and difficult to obtain in IIoT applications. In this paper,we shed light on the design of transfer learning (TL)-based systems for IIoT. We discuss how TL can overcome the demand for high-quality large data samples required to training ML models in IIoT. We also highlight the work principles and daunting challenges faced during the TL systems for IIoT. Furthermore,we categorize the TL systems for IIoT into TL for IIoT machinery level and for IIoT networking level and provide an in-depth discussion of the design building blocks and challenges of TL systems in each proposed class. Finally,we point out some future research directions for the design of novel TL-based systems for envisioned 5G-enabled IIoT applications.

With increasing usage of face recognition algorithms, it is well established that external artifacts and makeup accessories can be applied to different facial features such as eyes, nose, mouth, and cheek, to obfuscate one’s identity or to impersonate someone else’s identity. Recognizing faces in the presence of these artifacts comprises the problem of disguised face recognition, which is one of the most arduous covariates of face recognition. The challenge becomes exacerbated when disguised faces are captured in real-time environment, with low resolution images. To address the challenge of disguised face recognition, this paper first proposes a novel multi-objective encoder-decoder network, termed as DED-Net. DED-Net attempts to learn the class variations in the feature space generated by both disguised as well non-disguised images, using a combination of Mahalanobis and Cosine distance metrics, along with Mutual Information based supervision. The DED-Net is then extended to learn from the local and global features of both disguised and non-disguised face images for efficient face recognition, and the complete framework is termed as Disguise Resilient (D-Res) framework. The efficacy of the proposed framework has been demonstrated on two real-world benchmark datasets: Disguised Faces in the Wild (DFW) 2018 and DFW2019 competition datasets. In addition, this research also emphasizes on the importance of recognizing disguised faces in low resolution settings and proposes three experimental protocols to simulate the real-world surveillance scenario. To this effect, benchmark results have been shown on seven protocols for three low resolution settings (32×32, 24×24, and 16×16) of the two DFW benchmark datasets. The results demonstrate superior performance of the D-Res framework, in comparison with benchmark algorithms. For example, an improvement of around 3% is observed on the Overall protocol of the DFW2019 dataset, where the D-Res framework achieves 96.3%. Experiments have also been performed on benchmark face verification datasets (LFW, YTF, and IJB-B), where the D-Res framework achieves improved verification accuracy.

Soft gripping provides the potential for high performance in challenging tasks through morphological computing; however, design explorations are limited by a combination of a difficulty in generating useful models and use of laborious fabrication techniques. We focus on a class of grippers based on granular jamming that are particularly difficult to model and introduce a "one shot" technique that exploits multimaterial three-dimensional (3D) printing to create entire grippers, including membrane and grains, in a single print run. This technique fully supports the de facto physical generate-and-test methodology used for this class of grippers, as entire design iterations can be fitted onto a single print bed and fabricated from Computer-Aided Design (CAD) files in a matter of hours. Initial results demonstrate the approach by rapidly prototyping in materio solutions for two challenging problems in unconventional design spaces; a twisting gripper that uses programmed deformations to reliably pick a coin, and a multifunctional legged robot paw that offers the ability for compliant locomotion over rough terrains, as well as being able to pick objects in cluttered natural environments. The technique also allows us to easily characterize the design space of multimaterial printed jamming grippers and provide some useful design rules. The simplicity of our technique encourages and facilitates creativity and innovation. As such, we see our approach as an enabling tool to make informed principled forays into unconventional design spaces and support the creation of a new breed of novel soft actuators.

The emergence of wearable devices has opened up new potentials for egocentric activity recognition. Although some methods integrate attention mechanisms into deep neural networks to capture fine-grained human-object interactions in a weak-supervision manner, they either ignore exploiting the temporal consistency or generate attention based on considering appearance cues only. To address these limitations, in this paper, we propose an enhanced attention-tracking method, combined with multi-branch network (EAT-MBNet), for egocentric activity recognition. Specifically, we propose class-aware attention maps (CAAMs) by employing a self-attention-based module to refine the class activation maps (CAMs). Our proposed method can enhance the semantic dependency between the activity categories and the feature maps. To highlight the discriminative features from the regions of interest across frames, we propose a flow-guided attention-tracking (F-AT) module, by simultaneously leveraging historical attention and motion patterns. Furthermore, we propose a cross-modality modeling branch based on an inter-active GRU module, which captures the time-synchronized long-term relationships between the appearance and motion branches. Experimental results on four egocentric activity benchmarks demonstrate that the proposed method achieves state-of-the-art performance.

New class detection and effective model expansion are of great importance in incremental data mining. In open incremental data environments, data often come with novel classes, e.g., the emergence of new classes in image classification or new topics in opinion monitoring, and is denoted as class-incremental learning (C-IL) in literature. There are two main challenges in C-IL: how to conduct novelty detection and how to update the model with few novel class instances. Most previous methods pay much attention to the former challenge while ignoring the problem of efficiently updating models. To solve this problem, we propose a novel framework to handle the incremental new class, named learning to classify with incremental new class (LC-INC), which can process these two challenges automatically in one unified framework. In detail, LC-INC utilizes a novel structure network to consider the prototype information between class centers of known classes and newly incoming instances, which can dynamically combine the prediction information with structure information to detect novel class instances efficiently. On the other hand, the proposed structure network can also act as a meta-network, which can learn to expand the model much faster and more efficiently with inadequate novel class instances. Experiments on synthetic and real-world datasets successfully validate the effectiveness of our proposed method.

Multi-label learning deals with training examples each represented by a single instance while associated with multiple class labels. Due to the exponential number of possible label sets to be considered by the predictive model, it is commonly assumed that label correlations should be well exploited to design an effective multi-label learning approach. On the other hand, class-imbalance stands as an intrinsic property of multi-label data which significantly affects the generalization performance of the multi-label predictive model. For each class label, the number of training examples with positive labeling assignment is generally much less than those with negative labeling assignment. To deal with the class-imbalance issue for multi-label learning, a simple yet effective class-imbalance aware learning strategy called cross-coupling aggregation (COCOA) is proposed in this article. Specifically, COCOA works by leveraging the exploitation of label correlations as well as the exploration of class-imbalance simultaneously. For each class label, a number of multiclass imbalance learners are induced by randomly coupling with other labels, whose predictions on the unseen instance are aggregated to determine the corresponding labeling relevancy. Extensive experiments on 18 benchmark datasets clearly validate the effectiveness of COCOA against state-of-the-art multi-label learning approaches especially in terms of imbalance-specific evaluation metrics.

In recent years, many interpretable methods based on class activation maps (CAMs) have served as an important judging basis for the predictions of convolutional neural networks (CNNs). However, these methods still suffer from the problems of gradient noise, weight distortion, and perturbation deviation. In this work, we present self-attention class activation map (SA-CAM) and shed light on how it uses the self-attention mechanism to refine the existing CAM methods. In addition to generating basic activation feature maps, SA-CAM adds an attention skip connection as a regularization item for each feature map which further refines the focus area of an underlying CNN model. By introducing an attention branch and constructing a new attention operator, SA-CAM greatly alleviates the limitations of the CAM methods. The experimental results on the ImageNet dataset show that SA-CAM can not only generate highly accurate and intuitive interpretation but also have robust stability in adversarial comparison with the state-of-the-art CAM methods.

With the rise of custom silicon chips for AI acceleration, fair and comprehensive benchmarking of hardware innovations has become increasingly important. While benchmarking at the application- and system-level provides the most complete picture of trade-offs across multiple design dimensions, this can hide the impact of innovations at lower levels. Moreover, system-level benchmarking is not always feasible, especially for academic and industrial research chips. Benchmarking machine learning chips at lower abstraction levels is therefore still useful and common practice, making use of architecture- or circuit-level metrics. Yet the selection of good metrics and benchmarking conditions is critical, as these strongly influence the correlation between the observed performance benefits and the final system-level benefits. This paper provides an overview of benchmarking strategies at different abstraction levels, and discusses best practices and pitfalls to-be-avoided. We then propose to combine this diligent benchmarking at the appropriate individual abstraction level, with careful extrapolation to the system-level to also gauge impact at that topmost level. While the paper focuses on neural network inference workloads, many guidelines discussed here are applicable to the broader class of machine learning chips.

In this paper, we propose the co-attention dictionary network (CODNet) for weakly-supervised semantic segmentation using only image-level class labels. The CODNet model exploits extra semantic information by jointly leveraging a pair of samples with common semantics through co-attention rather than processing them independently. The inter-sample similarities of spatially distributed deep features are computed to merge reference features through non-local connections. To discover similar patterns regardless of appearance variations, we propose to extract image representations by equipping the neural networks with dictionary learning which provides the universal basis elements for different images. Based on the CODNet model, we propose a multi-reference class activation map (MR-CAM) algorithm which generates semantic segmentation masks for a target image by jointly merging semantic cues from multiple reference images. Experimental results on the PASCAL VOC 2012 and MSCOCO benchmark datasets for weakly-supervised semantic segmentation show that the proposed algorithm performs favorably against the state-of-the-art methods.

This note considers the design and analysis of a distributed sensor fault accommodation scheme for a class of nonlinear dynamical systems, modelled as interconnected subsystems. Each subsystem is subject to possible local sensor faults or may be impacted by a sensor fault in a neighboring subsystem due to the exchange of information between neighboring controllers. The proposed adaptive approximation-based distributed fault accommodation scheme is triggered in the event that a sensor fault is detected. The stability of the closed-loop system in the presence of sensor faults is rigorously analyzed both for the period prior to the detection of the fault, as well as after the detection of the fault when the fault accommodation scheme is activated. The effectiveness of the proposed scheme is illustrated by a simulation example.

Social media sites are increasingly being adopted to support teaching practice in higher education. Learning Analytics (LA) dashboards can be used to reveal how students engage with course material and others in the class. However, research on the best practices of designing, developing, and evaluating such dashboards to support teaching and learning with social media has been limited. Considering the increasing use of Twitter for both formal and informal learning processes, this paper presents our design process and a LA prototype dashboard developed based on a comprehensive literature review and an online survey among 54 higher education instructors who have used Twitter in their teaching. Keywords : Learning analytics, teaching, dashboards, survey

We will provide food microbiology laboratory training for all students taking this class and emphasize the importance of lab notebook record, and finally we will show students the examples of good lab notebook record.

This study presents a method for predicting location classes of a room such as a kitchen, and restroom, where a user is located by discovering location-specific sensor data motifs in sensor data observed by user’s sensor devices, such as smartwatch, without requiring labeled training data collected in a target environment. For example, we can observe similar waveforms corresponding to kitchen knife chopping actions using body-worn accelerometers in kitchens and can also observe similar sound features by active sound probing in bathrooms because of their water-resistant walls. This indicates that such location-specific sensor data motifs can be inherent information for location class prediction in almost every environment. This study proposes a novel method that automatically detects location-specific motifs from time series sensor data by calculating a score that represents the “location specificity” of each motif in a time series. Previous studies on location class prediction assume that location-specific sensor data are always observed in a room or use handcrafted rules and templates to detect location-specific sensor data resulting in difficulties in applying them to several realistic environments. In contrast, our method, named IndoLabel, can automatically discover short sensor data motifs, specific to a location class, and can automatically build an environment-independent location classifier without requiring handcrafted rules and templates. The proposed method was evaluated in real house environments using leave-one-environment-out cross-validation and achieved a state-of-the-art performance although labeled training data in the target environment was unavailable.

We present a novel model called One Class Minimum Spanning Tree (OCmst) for novelty detection problem that uses a Convolutional Neural Network (CNN) as deep feature extractor and graph-based model based on Minimum Spanning Tree (MST). In a novelty detection scenario, the training data is no polluted by outliers (abnormal class) and the goal is to recognize if a test instance belongs to the normal class or to the abnormal class. Our approach uses the deep features from CNN to feed a pair of MSTs built starting from each test instance. To cut down the computational time we use a parameter $\gamma$ to specify the size of the MST's starting to the neighbours from the test instance. To prove the effectiveness of the proposed approach we conducted experiments on two publicly available datasets, well-known in literature and we achieved the state-of-the-art results on CIFAR10 dataset., Comment: 16 pages

In this paper, the issues of both fault estimation and accommodation are studied for a class of continuous-time Markov jump linear systems under actuator fault, sensor fault and external disturbance in the framework of finite frequency domain. The imprecise statistic of modes transitions are considered here, which means that the transition rates are uncertain. Firstly, the sensor fault is defined as a new state, and then an adaptive observer is developed to estimate the actuator fault and the newly defined state with the aid of the descriptor system approach. By using the obtained fault estimation, a novel fault accommodation scheme is proposed based on the sample point controller design approach to satisfy the stochastic stability requirement of the closed-loop system with a prescribed H ∞ performance bound. Moreover, for the convenience of design, the presented sample point controller design is successfully transformed into resolving a class of input-delay control issues. Finally, two numerical examples, including a practical F-404 aircraft engine system, are illustrated to show the effectiveness and applicability of the developed design methods.

Cooperative problems under continuous control have always been the focus of multi-agent reinforcement learning. Existing algorithms suffer from the problem of uneven learning degree with the increase of the number of agents. In this paper, a new structure for a multi-agent actor critic is proposed, and the self-attention mechanism is applied in the critic network and the value decomposition method used to solve the uneven problem. The proposed algorithm makes full use of the samples in the replay memory buffer to learn the behavior of a class of agents. First, a new update method is proposed for policy networks that promotes learning efficiency. Second, the utilization of samples is improved, at the same time reflecting the ability of perspective-taking among groups. Finally, the ”deceptive signal” in training is eliminated and the learning degree among agents is more uniform than in the existing methods. Multiple experiments were conducted in two typical scenarios of a multi-agent particle environment. Experimental results show that the proposed algorithm can perform better than the state-of-the-art ones, and that it exhibits higher learning efficiency with an increasing number of agents.

The article analyzes foreign research and world experience in the fi eld of military robotics; analysis of approaches to the organization of development of groundbased robotic systems of heavy and middle class, taking into account the economic and technological development of the developing countries; assessment of the state of development and comparative analysis of their main technical characteristics. The analysis of modern armed confl icts allowed us to determine the most important combat tasks of groundbased robotic systems of heavy and medium class and their conceptual outlines. The analysis of the main technical characteristics and design features of domestic samples is carried out. They were developed by Ukrainian enterprises for their own working capital on a volunteer basis and have been repeatedly demonstrated at international exhibitions of weapons. It is noted that in Ukraine today there are no systematic approaches to creating multifunctional ground-based robotic systems for ground forces. There is no strategy for their use in advanced combat operations for performing special tasks and military operations. It is suggested that the basis of strategic planning for the use of ground-based robotic systems and planning for equipping the Ukrainian armed forces with them should be a political decision and a program for reforming the Ukrainian armed forces, which will indicate the place and role of ground-based robotic systems in performing combat and special tasks. Based on the analysis, a list of critical technologies necessary for creating domestic ground-based robotic systems of heavy and medium class at the level of foreign counterparts has been developed. The main components of certain critical technologies are described in detail and short-term, medium-term and long-term horizons for their development are defi ned. The article suggests the main approaches to creating such domestic complexes and also indicates the main eff orts that are advisable to concentrate to eliminate the lag behind the world leaders in the fi eld of creating military robots. Recommendations are given for the implementation of the necessary priority organizational measures and system steps to create ground-based robotic systems of heavy and medium classes in special conditions. The directions of further research are defined.

Distributed denial of service (DDoS) attacks remain challenging to mitigate in existing systems, including in-home networks that comprise different Internet of Things (IoT) devices. In this paper, we present a DDoS traffic detection model that uses a boosting method of logistic model trees for different IoT device classes. Specifically, a different version of the model will be generated and applied for each device class, since the characteristics of the network traffic from each device class may have subtle variation(s). As a case study, we explain how devices in a typical smart home environment can be categorized into four different classes (and in our context, Class 1 -very high level of traffic predictability, Class 2 -high level of traffic predictability, Class 3 -medium level of traffic predictability, and Class 4 -low level of traffic predictability). Findings from our evaluations show that the accuracy of our proposed approach is between 99.92% and 99.99% for these four device classes. In other words, we demonstrate that we can use device classes to help us more effectively detect DDoS traffic.

Imbalanced data can always be observed in our daily life and various practical tasks. A lot of well-constructed machine learning methodologies may produce ineffective performance, when conducted on this kind of data. This originates from the produced high training biases that towards the majority class instances. Among all the solutions of this problem, data generation of the minority class is always considered the most effective approach. However, in all the previous works, data are always processed sample-wisely and the distribution of each single data attribute is never noticed. So, in this paper, to estimate the mechanism of how each attribute contributes to its label, we explore the potential connection between the two items by Conditional Generative Adversarial Networks (CGAN) separately and individually. Then, the constructed new instances are purified by a designed attribute-based minimax filter and the survivors are concatenated to form the eventual generated data. In other words, different from the CGAN based data generation way, the proposed approach improves it by additionally considering all the single attribute patterns of the data that to construct new instances. In addition, we extend the binary class imbalanced learning framework to multiple class one. In the experimental part, the improved model is compared against GAN, CGAN and some other standard multiple-class oversampling algorithms on several widely used datasets. Results, in terms of four common measurements, have shown that the proposed approach can produce comparable and always superior performance when compared with the competitors.

The greater antihypertensive responses to initial therapy with calcium channel blockers (CCBs) or thiazide-type diuretics than renin-angiotensin system blockers as initial therapy in non-Hispanic Black (NHB) adults was recognized in the US High BP guidelines from 1988 to 2003. The 2014 Report from Panel Members Appointed to the Eighth Joint National Committee (2014 aJNC8 Report) and the 2017 American College of Cardiology/American Heart Association High Blood Pressure Guideline were the first to recommend CCBs or thiazide-type diuretics rather than renin-angiotensin system blockers as initial therapy in NHB. We assessed the temporal relationship of these recommendations on self-reported CCB or thiazide-type diuretics monotherapy by NHB and NHW adults with hypertension absent compelling indications for β-blockers or renin-angiotensin system blockers in National Health and Nutrition Examination Surveys 2015 to 2018 versus 2007 to 2012 (after versus before 2014 aJNC8 Report). CCB or thiazide-type diuretics monotherapy was unchanged in NHW adults (17.1% versus 18.1%, P =0.711) and insignificantly higher after 2014 among NHB adults (43.7% versus 38.2%, P =0.204), although CCB monotherapy increased (29.5% versus 21.0%, P =0.021) and renin-angiotensin system blocker monotherapy fell (44.5% versus 31.0%, P =0.008). Although evidence-based CCB monotherapy increased among NHB adults in 2015 to 2018, hypertension control declined as untreated hypertension and monotherapy increased. While a gap between recommended and actual monotherapy persists, evidence-based monotherapy appears insufficient to improve hypertension control in NHB adults, especially given evidence for worsening therapeutic inertia. Initiating treatment with single-pill combinations and timely therapeutic intensification when required to control hypertension are evidence-based, race-neutral options for improving hypertension control among NHB adults.

In people's daily lives, travel takes up an important part, and many trips are generated everyday, such as going to school or shopping. With the widely adoption of GPS-integrated devices, a large amount of trips can be recorded with GPS trajectories. These trajectories are represented by sequences of geo-coordinates and can help us answer simple questions such as “where did you go”. However, there is another important question awaiting to be answered, that is “what did/will you do”, i.e., the trip purpose inference. In practice, people's trip purposes are very important in understanding travel behaviors and estimating travel demands. Obviously, it is very challenging to infer trip purposes solely based on the trajectories, because the GPS devices are not accurate enough to pinpoint the venues visited. In this paper, we infer individual's trip purposes by combining the knowledge from heterogeneous data sources including trajectories, POIs and social media data. The proposed Dynamic Bayesian Network model (DBN) captures three important factors: the sequential properties of trip activities, the functionality and POI popularity of trip end areas. In addition, we propose an efficient method with local candidate pools to identify POIs from geo-tagged social media messages, and learn the POI popularities from nearby social media data. Moreover, trip data is usually imbalanced across different activities. This data imbalance problem can cause serious challenges because the DBN model could be biased by those “popular” class labels. Considering this challenge, we propose an ensemble DBN method with sampling technique (eDBN) which results in more accurate inference. Furthermore, real-world trip data are continuously collected on a daily basis. The batch model would result in unnecessary computation because historical data need to be revisited. We handle this problem by proposing an incremental DBN method (iDBN) which is both effective and efficient. Extensive experiments are conducted on real-world data sets with trajectories of 8,361 residents and the 6.9 million geo-tagged tweets in the Bay area. Experimental results demonstrate the advantages of the proposed method on correctly inferring the trip purposes.

In the fields of deep learning and computer vision,the security of object detection models has received extensive attention. Revealing the security vulnerabilities resulting from adversarial attacks has become one of the most important research directions. Existing studies show that object detection models can also be threatened by adversarial examples,just like other deep neural networks based models,e.g.,those for classification. In this paper,we propose a bidirectional adversarial attack method. Firstly,the added perturbation pushes the prediction results given by the object detectors far away from the groundtruth class while getting close to the background class. Secondly,a condence loss function is designed for the region proposal network to reduce the foreground scores. Thirdly,the adversarial examples are generated by a pre-trained autoencoder,and the model is trained using an adversarial approach,which can enhance the similarity between the adversarial examples and the original image and speed up algorithm convergence. The proposed method was verified on the most popular two-stage detection framework (Faster R-CNN),and 55.1% mAP-drop were obtained. In addition,the adversarial examples have superior transferability,migrating which to the common one-stage detection framework (YOLOv3) gets a 39.5% mAP-drop.

In the past twenty years, dozens of collaboration-based languages have emerged. Often they have an abstraction to denote a collaboration. Most of them use a form of single inheritance to build collaboration from another one. In this model, when a collaboration uses another one, every role class in the sub collaboration inherits from a role class in the super collaboration which has the same name. This affects the reusability of roles and collaborations and makes them semi interactive. To overcome this problem, this paper uses a new collaboration-based language named Set Oriented Programming (SOP). In SOP, collaborations, instead of being inherited, are composed with each other. To improve the reusability of SOP roles and to make them interactive, events are used. Events cause interaction of roles by crosscutting them. Moreover, a new context problem is discussed and dynamic context solution is proposed by delegations. Finally, we show that events improve the reusability and delegations concur context problem.

The digital revolution we are witnessing nowadays goes hand in hand with a revolution in cybercrime. This irrefutable fact has been a major reason for making digital forensic (DF) a pressing and timely topic to investigate. Thanks to the file system which is a rich source of digital evidence that may prove or deny a digital crime. Yet, although there are many tools that can be used to extract potentially conclusive evidence from the file system, there is still a need to develop effective techniques for evaluating the extracted evidence and link it directly to a digital crime. Machine learning can be posed as a possible solution looming in the horizon. This article proposes an Enhanced Multiclass Support Vector Machine (EMSVM) model that aims to improve the classification performance. The EMSVM suggests a new technique in selecting the most effective set of parameters when building a SVM model. In addition, since the DF is considered a multiclass classification problem duo to the fact that a file system might be accecced by more than one application, the EMSVM enhances the class assignment mechanism by supporting multi-class classification. The article then investigates the applicability of the proposed model in analysing incriminating digital evidence by inspecting the historical activities of file systems to realize if a malicious program manipulated them. The results obtained from the proposed model were promising when compared to several machine-learning algorithms.

Augmented reality is a device in combining objects between 2D and 3D virtual worlds, there's information that has not been implemented directly to the user, then by responding into scope that is 3D on the object with the resulting time at the same time to read in real-time, related to learning media where the teaching and learning process generally uses a book, then develops using projectors and laptops as information technology, besides the development of mobile (cellular) or android-based information technology that can make it easier for learning, especially the introduction of the solar system, the beginning of the solar system introduced various space objects both planets and surrounding space objects, in class VI who received the material as a basis before going to the next level, but the teaching and learning process there are several factors that influence i if it is examined that there are elderly teachers who have difficulty keeping up with technological developments, from learning media support tools such as projectors and teaching aids are still very limited, and not all teachers use information technology as a learning medium, the purpose of which is to make an introduction to AR learning media applications android-based solar system for class VI, based on these 3 problems as a solution an educational application is needed that supports learning activities using android augmented reality for the introduction of the solar system with the marker based tracking method, becomes a marker of 3D objects to bring up from the virtual world to the real which contains information on each of these objects, in other words visualizing the results of the marker directly into the real world, the final result is the creation of an AR application for learning the introduction of the solar system.

Applicable human-robot collaboration requires intuitive recognition of human intention during shared work. A grasped object such as a tool held by the human provides vital information about the upcoming task. In this letter, we explore the use of a wearable device to non-visually recognize objects within the human hand in various possible grasps. The device is based on Force-Myography (FMG) where simple and affordable force sensors measure perturbations of forearm muscles. We propose a novel Deep Neural-Network architecture termed Flip-U-Net inspired by the familiar U-Net architecture used for image segmentation. The Flip-U-Net is trained over data collected from several human participants and with multiple objects of each class. Data is collected while manipulating the objects between different grasps and arm postures. The data is also pre-processed with data augmentation and used to train a Variational Autoencoder for dimensionality reduction mapping. While prior work did not provide a transferable FMG-based model, we show that the proposed network can classify objects grasped by multiple new users without additional training efforts. Experiment with 12 test participants show classification accuracy of approximately 95% over multiple grasps and objects. Correlations between accuracy and various anthropometric measures are also presented. Furthermore, we show that the model can be fine-tuned to a particular user based on an anthropometric measure.

Background and Objective: The Flipped Classroom (FC) approach has become increasingly predominant and popular in medical education. This study aimed to explore the usefulness and the scope of FC based on medical students’ experience, with their adaptation challenges. Methods: The present study was a mixed-method accomplished during the academic years 2019-20, involving fourth-year students at the College of Medicine in Riyadh, Saudi Arabia. A self-administered questionnaire was used to seek their first experience and opinion of the FC. Results: A total of 234 questionnaires were distributed to the students, and 214 students completed the survey (response rate of 91.45%). Out of this total, 68.2 % were males and 31.8% were females. Most of the students agreed 156 (72.9%) that the flipped classroom was more engaging than the traditional lecture, among them 100 (68.5%) males and 56 (82.3) females agreed. Almost ~79% of students liked FC as it enabled them knowing the material in advance, and the class time was spent clarifying the facts and principles with active interaction, as commented during focus group discussion “More chance for discussing with the doctors, and I got the chance to answer” (St. 6). Conclusion: The results showed that the students like the FC more than the conventional classroom. Suggestions were given by students to improve the active learning sessions within the FC modality. doi: https://doi.org/10.12669/pjms.38.1.4256 How to cite this:Abdulghani HM, Sattar K, Ahmad T, Halepoto DM, Akram A. An insightful estimation of undergraduate medical students’ experience about the Flipped Classroom. Pak J Med Sci. 2022;38(1):69-75. doi: https://doi.org/10.12669/pjms.38.1.4256 This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Credit scoring models are usually developed using logistic regression. For several reasons, professionals of this area frequently categorize the quantitative covariates before using them in the model. In this work, we introduce a class of methods for covariate categorization in regression models for binary response variables. Applications to real data and a Monte Carlo simulation study suggest that one of the methods of this class has a better predictive performance and a smaller computational cost than other methods available in the literature.

The increasing use of machine learning in practice and legal regulations like EU’s GDPR cause the necessity to be able to explain the prediction and behavior of machine learning models. A prominent example of particularly intuitive explanations of AI models in the context of decision making are counterfactual explanations. Yet, it is still an open research problem how to efficiently compute counterfactual explanations for many models. In this contribution, we investigate how to efficiently compute counterfactual explanations for an important class of models, prototype-based classifiers such as learning vector quantization models. In particular, we derive specific convex and non-convex programs depending on the used metric. Typically counterfactual explanations deliver a feedback in terms of changes of the input features which lead to a different output – one application scenario is the link of these required changes to actionable items to change the desired outcome. Yet, rather than minimum changes of the input, it is interesting to address minimum changes of the model itself, which are required to lead to a different result. rather than a change of its inputs. We phrase this question as a counterfactual of the model prescription rather than the data points. We focus on distance-based classifiers (in particular learning vector quantization models), where model changes correspond to changes of metric parameters, and we develop efficient optimization techniques to generate such counterfactual metric changes depending on the chosen model.

This work shows the improvement in comprehending counter-current liquid-liquid extraction by applying jigsaw-type cooperative learning and the engineering math software Mathcad® in a chemical engineering course, part of the Chemistry degree. This study was performed on two different groups at the University of Almeria (Spain) over three academic years. The students were divided into two groups: one half of the class followed a non-cooperative learning methodology (the control) while the other half were spread among the jigsaw cooperative groups following the methodology known as “Jigsaw Experts Groups”. A main template made with Mathcad® of a multistage counter-current liquid-liquid extraction was supplied to both the jigsaw and non-jigsaw groups. The assessment of this educational experience in the course revealed that the jigsaw group outperformed the control group. The use of Mathcad® proved to be very intuitive and effective in explaining these relatively complex problems and utilising it is highly recommended; we suggest it is used rather than classical and less intuitive graphical methods.