Object detection is an important part of remote sensing image analysis to accurately identify key features in Earth observation data while minimizing resource usage. However, challenges such as multi-scale object representation and the identification of small-sized objects, particularly in compressed images, persist. To tackle this problem, we propose a state-of-the-art detector called Compressed Multiscale Feature Fusion YOLOv8 (CMSFF-YOLOv8) that employs a four-channel DWT coefficient image representation for preprocessing. This method processes the four coefficients of the 3-DWT result, arranging them into a 4-channel stacked image representation that retains the important spatial-frequency characteristics of the information. The improved preprocessing architecture design employs frequency domain filtering to retain important edge information, adaptive histogram equalization to adjust the intensity distribution locally, and multi-scale feature transfer to enable the extraction of important edge features at different scales. Squeeze-and-Excitation Convolution (SEConv) for channel-wise attention and Multi-Granularity Enhanced Feature Aggregation (MGEFA) for multi-level feature extraction of DWT coefficients are integrated into the backbone of the CMSFF-YOLOv8 architecture, enabling the selective prioritization of structural and edge features in complex settings. Spatial Pyramid Pooling Fast (SPPF) combined with shallow feature fusion networks (SFFNs) enhances small object detection. A custom neck design with upsampling facilitates multi-scale feature fusion. The result is mapped back to the LL3 channel in the compressed domain, reducing computational complexity while retaining critical global context. The proposed CMSFF-YOLOv8 detector, combined with a four-channel image representation and advanced preprocessing, demonstrates an efficient and accurate solution for object identification in compressed-domain remote sensing applications, achieving an F1-score of 0.8473, a recall of 0.82165, a precision of 0.8746, and mAP values of 0.90895 at IoU = 0.5 and 0.60063 at IoU = 0.75 on the HRPlanesV2 dataset.

With the widespread adoption of virtualization technology, it is imperative to strengthen its security, and dynamically modeling and instantly trapping malicious behaviors are challenging problems. Extant detection methods will be invalidated after the evasive malware manipulates the behavior trace. Currently, there is no approach to model the complex dynamic behavior of evasive malware, leading to missed opportunities for optimal detection. This work first presents antibypassing four-stage dynamic behavior modeling for time-efficient evasive malware detection (AFDBM-TEMD). AFDBM-TEMD models the interaction between evasive malware and its execution environment, identifying the optimal detection phases for various evasive malware. Moreover, it traps the crucial instructions and system calls invoked by the evasive malware into the virtual machine monitor layer to obtain the dynamic behavior information (including transmitted parameters, execution time, process information, return values, etc.) to identify the malicious software. Experimental results show that AFDBM-TEMD achieves new state-of-the-art results, and the proposed dynamic behavior modeling method has wide applicability, while the average detection time reaches milliseconds. Specifically, the detection rate is improved from 0–56.52% to 100% in contrast with the comparative methods, and the detection speed is increased by more than six times.

In reading comprehension, generating sentence-level distractors is a significant task, which requires a deep understanding of the article and question. The traditional entity-centered methods can only generate word-level or phrase-level distractors. Although recently proposed neural-based methods like sequence-to-sequence (Seq2Seq) model show great potential in generating creative text, the previous neural methods for distractor generation ignore two important aspects. First, they didn't model the interactions between the article and question, making the generated distractors tend to be too general or not relevant to question context. Second, they didn't emphasize the relationship between the distractor and article, making the generated distractors not semantically relevant to the article and thus fail to form a set of meaningful options. To solve the first problem, we propose a co-attention enhanced hierarchical architecture to better capture the interactions between the article and question, thus guide the decoder to generate more coherent distractors. To alleviate the second problem, we add an additional semantic similarity loss to push the generated distractors more relevant to the article. Experimental results show that our model outperforms several strong baselines on automatic metrics, achieving state-of-the-art performance. Further human evaluation indicates that our generated distractors are more coherent and more educative compared with those distractors generated by baselines.

Many research efforts have been proposed from physical, networking, to application layers over Vehicular Ad hoc Networks (VANETs) to provide more safety and convenience to passengers. However, due to the highly dynamic topologies and frequent disconnections in VANET, various challenges are faced due to the use of IP that effects the data delivery and user experiences. Therefore, a new paradigm namely Information-Centric Networking (ICN) has been proposed aiming to replace the traditional Internet Protocol by using the content name as the pillar element and providing a distributed in-network caching to enhance the data dissemination & access, and reduce the network load & response latency. The use of ICN in a vehicular environment may require different caching placement strategies and replacement policies. To this end, we study, in this paper, the effectiveness of in-network caching for VANET, we simulate and compare various strategies in different scenarios. Furthermore, we provide different research guidelines to enhance the use of caching in such a challenging network.

Vehicular Named Data Network (VNDN) uses Named Data Network (NDN) as a communication enabler. The communication is achieved using the content name instead of the host address. NDN integrates content caching at the network level rather than the application level. Hence, the network becomes aware of content caching and delivering. The content caching is a fundamental element in VNDN communication. However, due to the limitations of the cache store, only the most used content should be cached while the less used should be evicted. Traditional caching replacement policies may not work efficiently in VNDN due to the large and diverse exchanged content. To solve this issue, we propose an efficient cache replacement policy that takes the quality of service into consideration. The idea consists of classifying the traffic into different classes, and split the cache store into a set of sub-cache stores according to the defined traffic classes with different storage capacities according to the network requirements. Each content is assigned a popularity-density value that balances the content popularity with its size. Content with the highest popularity-density value is cached while the lowest is evicted. Simulation results prove the efficiency of the proposed solution to enhance the overall network quality of service.

In contrast to the host-centric model where the communication is directed using the destination address, Information-Centric Networking (ICN) adopts the content name as the pillar network element to provide data discovery and delivery process, as well as in other network functionalities. Named Data Networking (NDN) is an active ICN project that uses hierarchical unbounded names. These names are used in both interest and data packets and other data structures that may consume more memory with long lookup time. This paper targets the naming aspect in vehicular named data networks and proposes a Name-to-Hash Encoding scheme. The idea consists of hashing each name components separately to a fixed length, then perform a heuristic Wu-Manber-like algorithm lookup process. The former process enhances the NDN to consume less memory compared to hierarchical names, the latter process provides a fast lookup time. We have evaluated the proposed scheme against different related solutions using real domain datasets. Both theoretical analysis and experiments prove that the proposed scheme is efficient in terms of complexity, memory consumption, and lookup time.

Information-Centric Networking (ICN) is a new communication paradigm that replaces the host addresses by the name of content; Named Data Networking (NDN) is a promising ICN architecture that has attracted research attention in recent years. NDN is a receiver-driven architecture implements pull-based communication in the form of one-interest-one-data. This model poses different challenges, especially from the transport layer perspective. In contact to IP-based networks where the congestion is handled in an end-to-end manner, NDN cannot apply the same concept, while most of the existing solutions are based on hop-by-hop connection. In this paper, we present a new congestion control mechanism for NDN based on link quality estimation. We focus our efforts to provide fast data transmission, decrease packet dropping rate, and maximize the link utilization. The simulation results show that our solution outperforms the NDN schemes in terms of throughput and drop packets.

Information-centric networking (ICN), a new networking paradigm in which the focal point is a named data, has been proposed recently as an evolving concept to the actual host-centric model of the Internet that relies mainly on host addresses. In vehicular networks, where vehicles are generally moving network elements and follow a content-oriented fashion, it will be fitting to use the ICN paradigm to improve the content dissemination and reduce the content retrieval latency. By applying this concept to such networks, we focus in this paper on the content delivery issue and propose an optimized caching scheme that proactively predicts the moving direction of a vehicle and brings into the next encountered RSU cache only the required content of interest to that vehicle. According to the obtained results from different measured metrics, the proposed solution outperforms in many ways other proposed schemes in the literature. For instance, our scheme improves drastically the cache utilization, enhances the network delay, and boosts the content diversity and distribution.

Mixture of regressions (MoR) is the well-established and effective approach to model discontinuous and heterogeneous data in regression problems. Existing MoR approaches assume smooth joint distribution for its good anlaytic properties. However, such assumption makes existing MoR very sensitive to intra-component outliers (the noisy training data residing in certain components) and the inter-component imbalance (the different amounts of training data in different components). In this paper, we make the earliest effort on Self-paced Learning (SPL) in MoR, i.e., Self-paced mixture of regressions (SPMoR) model. We propose a novel self-paced regularizer based on the Exclusive LASSO, which improves inter-component balance of training data. As a robust learning regime, SPL pursues confidence sample reasoning. To demonstrate the effectiveness of SPMoR, we conducted experiments on both the sythetic examples and real-world applications to age estimation and glucose estimation. The results show that SPMoR outperforms the state-of-the-arts methods.

Having a system to stratify individuals according to risk is key to clinical disease prevention. This allows individuals identified at different risk tiers to benefit from further investigation and intervention. But the same risk score estimated for two different persons does not mean they need the same further investigation or represent the similarity health condition between two persons. Meanwhile, users still do not know a prior what most of the risk tiers are, and how many tiers should be found in risk stratification. In this paper, the proposed pairwise and size constrained Kmeans (PSCKmeans) method simultaneously integrates the limited supervised information and the size constraints to screen the high-risk population based on similarity measurement, and gets a feasible and balanced stratification solution to avoid cluster with few points. Results on China Health and Nutrition Survey public dataset and follow-up dataset show that the proposed PSCKmeans method can naturally grade the risk of diabetes into four tiers, and achieve 73.8%, 85.1%, and 0.95% sensitivity, specificity, and ratio of minimum to expected on testing data. The proposed method compares favorably with eight previous semisupervised clustering methods; it demonstrates that semisupervised clustering by unifying multiple forms of constraints can guide a good partition that is more relevant for the domain and find new categories through prior knowledge. Finally, this risk stratification model can provide a tool for risk stratification of clinical disease and be used for further intervention for people with similar health condition.

Abstract The kernel‐level rootkits compromise the security of operating systems. In the current research studies, virtualization is used as a key tool against these attacks with virtualization‐based memory protection. There are glitches in the memory protection mechanism, and it is vulnerable to page mapping attack and hard to be used for protecting dynamic data. To address these problems, we proposed a secure paging mechanism and constructed an external and transparent architecture named multiple operating systems kernel guard (MOSKG), which can protect critical kernel data in different operating systems like Windows and Linux, both of 32‐bit and 64‐bit. To evaluate our proposed architecture, we applied some experiments that are based on the study of kernel rootkits. The results show that MOSKG can protect critical kernel data from dynamic kernel object manipulation and page mapping attack, and it defeats all of the kernel‐level attacks. It is also a significant conclusion that MOSKG only introduces a small performance overhead of 2.3%. Copyright © 2015 John Wiley & Sons, Ltd.

Diabetes mellitus is a chronic disease and a worldwide public health challenge. It has been shown that 50-80% proportion of T2DM is undiagnosed. In this paper, support vector machines are utilized to screen diabetes, and an ensemble learning module is added, which turns the "black box" of SVM decisions into comprehensible and transparent rules, and it is also useful for solving imbalance problem. Results on China Health and Nutrition Survey data show that the proposed ensemble learning method generates rule sets with weighted average precision 94.2% and weighted average recall 93.9% for all classes. Furthermore, the hybrid system can provide a tool for diagnosis of diabetes, and it supports a second opinion for lay users.