Medical imaging plays an indispensable role in precise patient diagnosis. The integration of deep learning into medical diagnostics is becoming increasingly common. However, existing deep learning models face performance and efficiency challenges, especially in resource-constrained scenarios. To overcome these challenges, we introduce a novel dendritic neural efficientnet model called DEN, inspired by the function of brain neurons, which efficiently extracts image features and enhances image classification performance. Assessments on a diabetic retinopathy fundus image dataset reveal DEN’s superior performance compared to EfficientNet and other classical neural network models.

This letter introduces an innovation for the heterogeneous storage architecture of AI chips, specifically focusing on the integration of six transistors(6T) and eight transistors(8T) hybrid SRAM. Traditional approaches to reducing SRAM power consumption typically involve lowering the operating voltage, a method that often substantially diminishes the recognition rate of neural networks. However, the innovative design detailed in this letter amalgamates the strengths of both SRAM types. It operates at a voltage lower than conventional SRAM, thereby significantly reducing the power consumption in neural networks without compromising performance.

We consider the problem of finding the best subset of sensors in wireless sensor networks where linear Bayesian parameter estimation is conducted from the selected measurements corrupted by correlated noise. We aim to directly minimize the estimation error which is manipulated by using the QR and LU factorizations. We derive an analytic result which expedites the sensor selection in a greedy manner. We also provide the complexity of the proposed algorithm in comparison with previous selection methods. We evaluate the performance through numerical experiments using random measurements under correlated noise and demonstrate a competitive estimation accuracy of the proposed algorithm with a reasonable increase in complexity as compared with the previous selection methods.

Chinese Spelling Correction (CSC) is an important natural language processing task. Existing methods for CSC mostly utilize BERT models, which select a character from a candidate list to correct errors in the sentence. World knowledge refers to structured information and relationships spanning a wide range of domains and subjects, while definition knowledge pertains to textual explanations or descriptions of specific words or concepts. Both forms of knowledge have the potential to enhance a model’s ability to comprehend contextual nuances. As BERT lacks sufficient guidance from world knowledge for error correction and existing models overlook the rich definition knowledge in Chinese dictionaries, the performance of spelling correction models is somewhat compromised. To address these issues, within the world knowledge network, this study injects world knowledge from knowledge graphs into the model to assist in correcting spelling errors caused by a lack of world knowledge. Additionally, the definition knowledge network in this model improves the error correction capability by utilizing the definitions from the Chinese dictionary through a comparative learning approach. Experimental results on the SIGHAN benchmark dataset validate the effectiveness of our approach.

Detecting key frames in videos has garnered substantial attention in recent years, it is a point-level task and has deep research value and application prospect in daily life. For instances, video surveillance system, video cover generation and highlight moment flashback all demands the technique of key frame detection. However, the task is beset by challenges such as the sparsity of key frame instances, imbalances between target frames and background frames, and the absence of post-processing method. In response to these problems, we introduce a novel and effective Temporal Interval Guided (TIG) framework to precisely localize specific frames. The framework is incorporated with a proposed Point-Level-Soft non-maximum suppression (PLS-NMS) post-processing algorithm which is suitable for point-level task, facilitated by the well-designed confidence score decay function. Furthermore, we propose a TIG-loss, exhibiting sensitivity to temporal interval from target frame, to optimize the two-stage framework. The proposed method can be broadly applied to key frame detection in video understanding, including action start detection and static video summarization. Extensive experimentation validates the efficacy of our approach on action start detection benchmark datasets: THUMOS’14 and Activitynet v1.3, and we have reached state-of-the-art performance. Competitive results are also demonstrated on SumMe and TVSum datasets for deep learning based static video summarization.

This work focuses on training dataset enhancement of informative relational triplets for Scene Graph Generation (SGG). Due to the lack of effective supervision, the current SGG model predictions perform poorly for informative relational triplets with inadequate training samples. Therefore, we propose two novel training dataset enhancement modules: Feature Space Triplet Augmentation (FSTA) and Soft Transfer. FSTA leverages a feature generator trained to generate representations of an object in relational triplets. The biased prediction based sampling in FSTA efficiently augments artificial triplets focusing on the challenging ones. In addition, we introduce Soft Transfer, which assigns soft predicate labels to general relational triplets to make more supervisions for informative predicate classes effectively. Experimental results show that integrating FSTA and Soft Transfer achieve high levels of both Recall and mean Recall in Visual Genome dataset. The mean of Recall and mean Recall is the highest among all the existing model-agnostic methods.

In this paper, the Towards High Performance Voxel-based 3D Object Detection (Voxel-RCNN) three-dimensional (3D) point cloud object detection model is used as the benchmark network. Aiming at the problems existing in the current mainstream 3D point cloud voxelization methods, such as the backbone and the lack of feature expression ability under the bird’s-eye view (BEV), a high-performance voxel-based 3D object detection network (Reinforced Voxel-RCNN) is proposed. Firstly, a 3D feature extraction module based on the integration of inverted residual convolutional network and weight normalization is designed on the 3D backbone. This module can not only well retain more point cloud feature information, enhance the information interaction between convolutional layers, but also improve the feature extraction ability of the backbone network. Secondly, a spatial feature-semantic fusion module based on spatial and channel attention is proposed from a BEV perspective. The mixed use of channel features and semantic features further improves the network’s ability to express point cloud features. In the comparison of experimental results on the public dataset KITTI, the experimental results of this paper are better than many voxel-based methods. Compared with the baseline network, the 3D average accuracy and BEV average accuracy on the three categories of Car, Cyclist, and Pedestrians are improved. Among them, in the 3D average accuracy, the improvement rate of Car category is 0.23%, Cyclist is 0.78%, and Pedestrians is 2.08%. In the context of BEV average accuracy, enhancements are observed: 0.32% for the Car category, 0.99% for Cyclist, and 2.38% for Pedestrians. The findings demonstrate that the algorithm enhancement introduced in this study effectively enhances the accuracy of target category detection.

In the realm of Single Image Super-Resolution (SISR), the meticulously crafted Nonlocal Sparse Attention-based block demonstrates its efficacy in noise reduction and computational cost reduction for nonlocal (global) features. However, it neglect the traditional Convolutional-based block, which proficient in handling local features. Thus, merging both the Nonlocal Sparse Attention-based block and the Convolutional-based block to concurrently manage local and nonlocal features poses a significant challenge. To tackle the aforementioned issues, this paper introduces the Channel Contrastive Attention-based Local-Nonlocal Mutual block (CCLN) for Super-Resolution (SR). (1) We introduce the CCLN block, encompassing the Local Sparse Convolutional-based block for local features and the Nonlocal Sparse Attention-based network block for nonlocal features. (2) We introduce Channel Contrastive Attention (CCA) blocks, incorporating Sparse Aggregation into Convolutional-based blocks. Additionally, we introduce a robust framework to fuse these two blocks, ensuring that each branch operates according to its respective strengths. (3) The CCLN block can seamlessly integrate into established network backbones like the Enhanced Deep Super-Resolution network (EDSR), achieving in the Channel Attention based Local-Nonlocal Mutual Network (CCLNN). Experimental results show that our CCLNN effectively leverages both local and nonlocal features, outperforming other state-of-the-art algorithms.

For flight simulators, it is crucial to create three-dimensional terrain using clear remote sensing images. However, due to haze and other contributing variables, the obtained remote sensing images typically have low contrast and blurry features. In order to build a flight simulator visual system, we propose a deep learning-based dehaze model for remote sensing images dehazing. An encoder-decoder architecture is proposed that consists of a multiscale fusion module and a gated large kernel convolutional attention module. This architecture can fuse multi-resolution global and local semantic features and can adaptively extract image features under complex terrain. The experimental results demonstrate that, with good generality and application, the model outperforms existing comparison techniques and achieves high-confidence dehazing in remote sensing images with a variety of haze concentrations, multi-complex terrains, and multi-spatial resolutions.

Automatic detection and recognition systems have numerous applications in smart city implementation. Despite the accuracy and widespread use of device-based and optical methods, several issues remain. These include device limitations, environmental limitations, and privacy concerns. The FMWC sensor can overcome these issues to detect and track moving people accurately in commercial environments. However, single-chip mmWave sensor solutions might struggle to recognize standing and sitting people due to the necessary static removal module. To address these issues, we propose a real-time indoor people detection and tracking fusion system using mmWave radar and cameras. The proposed fusion system approaches an overall detection accuracy of 93.8% with a median position error of 1.7 m in a commercial environment. Compared to our single-chip mmWave radar solution addressing an overall accuracy of 83.5% for walking people, it performs better in detecting individual stillness, which may feed the security needs in retail. This system visualizes customer information, including trajectories and the number of people. It helps commercial environments prevent crowds during the COVID-19 pandemic and analyze customer visiting patterns for efficient management and marketing. Powered by an IoT platform, the system can be deployed in the cloud for easy large-scale implementation.

Entity alignment (EA) is a crucial task for integrating cross-lingual and cross-domain knowledge graphs (KGs), which aims to discover entities referring to the same real-world object from different KGs. Most existing embedding-based methods generate aligning entity representation by mining the relevance of triple elements, paying little attention to triple indivisibility and entity role diversity. In this paper, a novel framework named TTEA - Type-enhanced Ensemble Triple Representation via Triple-aware Attention for Cross-lingual Entity Alignment is proposed to overcome the above shortcomings from the perspective of ensemble triple representation considering triple specificity and diversity features of entity role. Specifically, the ensemble triple representation is derived by regarding relation as information carrier between semantic and type spaces, and hence the noise influence during spatial transformation and information propagation can be smoothly controlled via specificity-aware triple attention. Moreover, the role diversity of triple elements is modeled via triple-aware entity enhancement in TTEA for EA-oriented entity representation. Extensive experiments on three real-world cross-lingual datasets demonstrate that our framework makes comparative results.

Unmanned Aerial Vehicle (UAV)-assisted Mobile Edge Computing (MEC) can provide mobile users (MU) with additional computing services and a wide range of connectivity. This paper investigates the joint optimization strategy of task offloading and resource allocation for UAV-assisted MEC systems in complex scenarios with the goal of reducing the total system cost, consisting of task execution latency and energy consumption. We adopt a game theoretic approach to model the interaction process between the MEC server and the MU Stackelberg bilayer game model. Then, the original problem with complex multi-constraints is transformed into a duality problem using the Lagrangian duality method. Furthermore, we prove that the modeled Stackelberg bilayer game has a unique Nash equilibrium solution. In order to obtain an approximate optimal solution to the proposed problem, we propose a two-stage alternating iteration (TASR) algorithm based on the subgradient method and the marginal revenue optimization method. We evaluate the effective performance of the proposed algorithm through detailed simulation experiments. The simulation results show that the proposed algorithm is superior and robust compared to other benchmark methods and can effectively reduce the task execution latency and total system cost in different scenarios.

Aiming at the problem that large-scale traffic data lack labels and take too long for feature extraction in network intrusion detection, an unsupervised intrusion detection method ACOPOD based on Adam asymmetric autoencoder and COPOD (Copula-Based Outlier Detection) algorithm is proposed. This method uses the Adam asymmetric autoencoder with a reduced structure to extract features from the network data and reduce the data dimension. Then, based on the Copula function, the joint probability distribution of all features is represented by the edge probability of each feature, and then the outliers are detected. Experiments on the published NSL-KDD dataset with six other traditional unsupervised anomaly detection methods show that ACOPOD achieves higher precision and has obvious advantages in running speed. Experiments on the real civil aviation air traffic management network dataset further prove that the method can effectively detect intrusion behavior in the real network environment, and the results are interpretable and helpful for attack source tracing.

Watermarking methods require robustness against various attacks. Conventional watermarking methods use error-correcting codes or spread spectrum to correct watermarking errors. Errors can also be reduced by embedding the watermark into the frequency domain and by using SIFT feature points. If the type and strength of the attack can be estimated, the errors can be further reduced. There are several types of attacks, such as scaling, rotation, and cropping, and it is necessary to aim for robustness against all of them. Focusing on the scaling tolerance of watermarks, we propose a watermarking method using SIFT feature points and DFT, and introduce a pilot signal. The proposed method estimates the scaling rate using the pilot signal in the form of a grid. When a stego-image is scaled, the grid interval of the pilot signal also changes, and the scaling rate can be estimated from the amount of change. The accuracy of estimating the scaling rate by the proposed method was evaluated in terms of the relative error of the scaling rate. The results show that the proposed method could reduce errors in the watermark by using the estimated scaling rate.

Software refactoring is an important process in software development. During software refactoring, code smell is a popular research topic that refers to design or implementation flaws in the software. Large class is one of the most concerning code smells in software refactoring. Detecting and refactoring such problem has a profound impact on software quality. In past years, software metrics and clustering techniques have commonly been used for the large class detection. However, deep-learning-based approaches have also received considerable attention in recent studies. In this study, we apply graph neural networks (GNNs), an important division of deep learning, to address the problem of large class detection. First, to support the extensive data requirements of the deep learning task, we apply a semiautomatic approach to generate a substantial number of data samples. Next, we design a new type of directed heterogeneous graph (DHG) as an input graph using the methods similarity matrix and software metrics. We construct an input graph for each class sample and make the graph classification with GNNs to identify the smelly classes. In our experiments, we apply three typical GNN model architectures for large class detection and compare the results with those of previous studies. The results show that the proposed approach can achieve more accurate and stable detection performance.

In this paper, we propose two algorithms, B-N2N and B-N2S, that solve the node-to-node and node-to-set disjoint paths problems in the bicube, respectively. We prove their correctness and that the time complexities of the B-N2N and B-N2S algorithms are O(n2) and O(n2 log n), respectively, if they are applied in an n-dimensional bicube with n ≥ 5. Also, we prove that the maximum lengths of the paths generated by B-N2N and B-N2S are both n + 2. Furthermore, we have shown that the algorithms can be applied in the locally twisted cube, too, with the same performance.

Emergency Medical Services (EMS) play a crucial role in healthcare systems, managing pre-hospital or out-of-hospital emergencies from the onset of an emergency call to the patient’s arrival at a healthcare facility. The design of an efficient ambulance location model is pivotal in enhancing survival rates, controlling morbidity, and preventing disability. Key factors in the classical models typically include travel time, demand zones, and the number of stations. While urban EMS systems have received extensive examination due to their centralized populations, rural areas pose distinct challenges. These include lower population density and longer response distances, contributing to a higher fatality rate due to sparse population distribution, limited EMS stations, and extended travel times. To address these challenges, we introduce a novel mathematical model that aims to optimize coverage and equity. A distinctive feature of our model is the integration of equity within the objective function, coupled with a focus on practical response time that includes the period required for personal protective equipment procedures, ensuring the model’s applicability and realism in emergency response scenarios. We tackle the proposed problem using a tailored genetic algorithm and propose a greedy algorithm for solution construction. The implementation of our tailored Genetic Algorithm promises efficient and effective EMS solutions, potentially enhancing emergency care and health outcomes in rural communities.

This article describes the idea of utilizing Attested Execution Secure Processors (AESPs) that fit into building a secure Self-Sovereign Identity (SSI) system satisfying Sybil-resistance under permissionless blockchains. Today’s circumstances requiring people to be more online have encouraged us to address digital identity preserving privacy. There is a momentum of research addressing SSI, and many researchers approach blockchain technology as a foundation. SSI brings natural persons various benefits such as owning controls; on the other side, digital identity systems in the real world require Sybil-resistance to comply with Anti-Money-Laundering (AML) and other needs. The main idea in our proposal is to utilize AESPs for three reasons: first is the use of attested execution capability along with tamper-resistance, which is a strong assumption; second is powerfulness and flexibility, allowing various open-source programs to be executed within a secure enclave, and the third is that equipping hardware-assisted security in mobile devices has become a norm. Rafael Pass et al.’s formal abstraction of AESPs and the ideal functionality $\color{brown}{\mathcal{G}_\mathtt{att}}$ enable us to formulate how hardware-assisted security works for secure digital identity systems preserving privacy under permissionless blockchains mathematically. Our proposal of the AESP-based SSI architecture and system protocols, $\color{blue}{\Pi^{\mathcal{G}_\mathtt{att}}}$, demonstrates the advantages of building a proper SSI system that satisfies the Sybil-resistant requirement. The protocols may eliminate the online distributed committee assumed in other research, such as CanDID, because of assuming AESPs; thus, $\color{blue}{\Pi^{\mathcal{G}_\mathtt{att}}}$ allows not to rely on multi-party computation (MPC), bringing drastic flexibility and efficiency compared with the existing SSI systems.

Radiation-induced temporal errors become a significant issue for circuit reliability. We measured the pulse widths of radiation-induced single event transients (SETs) from pMOSFETs and nMOSFETs separately. Test results show that heavy-ion induced SET rates of nMOSFETs were twice as high as those of pMOSFETs and that neutron-induced SETs occurred only in nMOSFETs. It was confirmed that the SET distribution from inverter chains can be estimated using the SET distribution from pMOSFETs and nMOSFETs by considering the difference in load capacitance of the measurement circuits.