In the cloud server, ranked searchable encryption allows the cloud server to search for the first k most relevant documents based on the correlation scores between the query keyword and the document. OPE (Orderpreserv...
详细信息
The rapid development of wireless communications have driven the need for careful optimization of network parameters to improve network performance and reduce operational cost. Traditional methods, however, struggle w...
详细信息
Spoofed noisy speeches seriously threaten the speech-based embedded systems, such as smartphones and intelligent assistants. Consequently, we present an anti-spoofing detection model with activation-based residual blo...
详细信息
Spoofed noisy speeches seriously threaten the speech-based embedded systems, such as smartphones and intelligent assistants. Consequently, we present an anti-spoofing detection model with activation-based residual blocks to identify spoofed noisy speeches with the requirements of high accuracy and low time overhead. Through theoretic analysis of noise propagation on shortcut connections of traditional residual blocks, we observe that different activation functions can help reducing the influence of noise under certain situations. Then, we propose a feature-aware activation function to weaken the influence of noise and enhance the anti-spoofing features on shortcut connections, in which a fine-grained processing is designed to remove noise and strengthen significant features. We also propose a variance-increasing-based optimization algorithm to find the optimal hyperparameters of the feature-aware activation function. Benchmark-based experiments demonstrate that the proposed method can reduce the average equal error rate of anti-spoofing detection from 21.72% to 4.51% and improve the accuracy by up to 37.06% and save up to 91.26% of time overhead on Jetson AGX Xavier compared with ten state-of-the-art methods.
Data confidentiality and availability in an open network environment are essential for Internet of Vehicles (IoV). Encryption technology can only ensure user privacy, but impedes the optimal use of vehicular data. As ...
详细信息
Data confidentiality and availability in an open network environment are essential for Internet of Vehicles (IoV). Encryption technology can only ensure user privacy, but impedes the optimal use of vehicular data. As a solution, public key encryption with equality test (PKEET) is proposed. However, most existing PKEET schemes rely on the homogeneity of communication parties and reliability of algorithms. The former requires both communication parties to be under the same cryptographic system, while the latter requires cloud server (CS) is a semi-honest entity. To address the challenge, we propose a heterogeneous signcryption scheme with equality test, named CP-HSCET, which allows CS to verify whether the plaintexts corresponding to ciphertexts are equal without unsigncrypting, thus enhancing the availability. It also enables an IoV device in certificateless (CLC) cryptosystem to signcrypt messages and send them to another device in public key infrastructure (PKI)based cryptosystem, thus achieving higher availability. To overcome the reliability problem, we incorporate blockchain technology into our algorithm, thus eliminating the dependence on an honest (or semi-honest) CS. We also prove its security notions and design it to satisfy confidentiality, integrity, non-repudiation, and authentication. Finally, we implement our scheme with Hyperledge Fabric and compare the proposed scheme with other four comparison schemes. The result shows that it reduces by 35.49%, 53.79%, 80.15% and 30.37% of other four schemes in terms of computation time at 500 ciphertexts. Therefore, our solution is the most suitable for IoV environments, where most of the devices have lower computational power.
Traditionally threat detection in organisations is reactive through pre-defined and preconfigured rules that are embedded in automated tools such as firewalls, anti-virus software, security information and event manag...
详细信息
ISBN:
(纸本)9781914587405
Traditionally threat detection in organisations is reactive through pre-defined and preconfigured rules that are embedded in automated tools such as firewalls, anti-virus software, security information and event management (SIEMs) and intrusion detection systems/intrusion prevention systems (IDS/IPS). As the fourth industrial revolution (4IR) brings with it an exponential increase in technological advances and global interconnectivity, the cyberspace presents security risks and threats the scale of which is unprecedented. These security risks and threats have the potential of exposing confidential information, damaging the reputation of credible organisations and/or inflicting harm. The regular occurrence and complexity of cyber intrusions makes the guarding enterprise and government networks a daunting task. Nation states and businesses need to be ingenious and consider innovative and proactive means of safeguarding their valuable assets. The growth of technological, physical and biological worlds necessitates the adoption of a proactive approach towards safeguarding cyber space. This paper centers on cyber threat hunting (CTH) as one such proactive and important measure that can be adopted. The paper has a central contention that effective CTH cannot be an autonomous ‘plug in’ or a standalone intervention. To be effective CTH has to be synergistically integrated with relevant existing fields and practices. Academic work on such conceptual integration of where CTH fits is scarce. Within the confines of the paper we do not attempt to integrate CTH with many of the various relevant fields and practices. Instead, we limit the scope to postulations on CTH’s interface with two fields of central importance in cyber security, namely Cyber Counterintelligence (CCI) and Cyber Threat Monitoring and Analysis (CTMA). The paper’s corresponding two primary objectives are to position CTH within the broader field of CCI and further contextualise CTH within the CTMA domain. The pos
The reuse and distribution of open-source software must be in compliance with its accompanying open-source license. In modern packaging ecosystems, maintaining such compliance is challenging because a package may have...
详细信息
This paper starts by revealing a surprising finding: without any learning, a randomly initialized CNN can localize objects surprisingly well. That is, a CNN has an inductive bias to naturally focus on objects, named a...
详细信息
This paper starts by revealing a surprising finding: without any learning, a randomly initialized CNN can localize objects surprisingly well. That is, a CNN has an inductive bias to naturally focus on objects, named as Tobias ("The object is at sight") in this paper. This empirical inductive bias is further theoretically analyzed and empirically verified, and successfully applied to self-supervised learning as well as supervised learning. For self-supervised learning, a CNN is encouraged to learn representations that focus on the foreground object, by transforming every image into various versions with different backgrounds, where the foreground and background separation is guided by Tobias. Experimental results show that the proposed Tobias significantly improves downstream tasks, especially for object detection. This paper also shows that Tobias has consistent improvements on training sets of different sizes, and is more resilient to changes in image augmentations. Furthermore, we apply Tobias to supervised image classification by letting the average pooling layer focus on foreground regions, which achieves improved performance on various benchmarks.
Nuclearmagnetic resonance imaging of breasts often presents complex *** tumors exhibit varying sizes,uneven intensity,and indistinct *** characteristics can lead to challenges such as low accuracy and incorrect segmen...
详细信息
Nuclearmagnetic resonance imaging of breasts often presents complex *** tumors exhibit varying sizes,uneven intensity,and indistinct *** characteristics can lead to challenges such as low accuracy and incorrect segmentation during tumor ***,we propose a two-stage breast tumor segmentation method leveraging multi-scale features and boundary attention ***,the breast region of interest is extracted to isolate the breast area from surrounding tissues and ***,we devise a fusion network incorporatingmulti-scale features and boundary attentionmechanisms for breast tumor *** incorporate multi-scale parallel dilated convolution modules into the network,enhancing its capability to segment tumors of various sizes through multi-scale convolution and novel fusion ***,attention and boundary detection modules are included to augment the network’s capacity to locate tumors by capturing nonlocal dependencies in both spatial and channel ***,a hybrid loss function with boundary weight is employed to address sample class imbalance issues and enhance the network’s boundary maintenance capability through additional *** was evaluated using breast data from 207 patients at RuijinHospital,resulting in a 6.64%increase in Dice similarity coefficient compared to the *** results demonstrate the superiority of the method over other segmentation techniques,with fewer model parameters.
As an important indicator of economic development, house price prediction is a hot issue in the field of business and research in urban computing. However, due to a variety of influencing factors and sparse transactio...
详细信息
The classification of short-term power load data by clustering algorithm can lay a good foundation for the subsequent power load forecasting work and provide a more efficient, safe and reliable direction for the opera...
详细信息
暂无评论