Compared with a single smart contract, interactive smart contracts have more complex interactions such as mutual calling. However, existing smart contract detection and verification methods primarily consider the problems existing in a single smart contract, and the correctness of interactive smart contracts cannot be guaranteed. To address this limitation, this paper proposes a method for verifying the correctness of interactive smart contracts, where behavior interaction priority (BIP) modeling is built for interactive smart contract systems and Solidity deployment diagram (SDD) is introduced to describe contract interactions. By employing formal verification techniques, the proposed approach ensures the correctness of the interactive smart contracts and realizes their reconstruction. Experimental results show that the proposed method can effectively verify the correctness of interactive contract systems.

UAV-assisted communication system is an important component of future wireless networks. In order to further improve the utilization of time-frequency resources in UAV-assisted communication systems, this paper proposes a communication architecture based on non-orthogonal multiple access (NOMA) technology and introduces a TD3-TOPATM (twin delayed-trajectory optimization and power allocation for total throughput maximization) algorithm based on the double-delay deep deterministic policy gradient strategy. The TD3-TOPATM algorithm jointly optimizes the 3D trajectory and power allocation strategy of the UAV, with the aim of maximizing the total throughput while satisfying constraints on maximum power, spatial boundaries maximum flight speed, and quality of service (QoS). Simulation results show that compared with the trajectory optimization algorithm with random optimization, the TD3-TOPATM algorithm achieves a performance gain of 98%. Additionally, it outperforms the deep Q-network (DQN)-based trajectory optimization and resource allocation algorithm, increasing total throughput by 19.4%, and surpasses the deep deterministic policy gradient (DDPG)-based algorithm with a 9.7% throughput gain. Furthermore, the NOMA-based UAV-assisted communication scheme achieves a 55% performance gain compared to the OMA-based scheme.

Collaborative filtering algorithms are widely used in recommendation systems, with a key research focus on how to achieve effective user clustering and identify more accurate sets of similar neighbors. One of the key research focuses is to achieve user clustering and identify more similar neighbor sets. To enhance the accuracy of classification and prediction in these algorithms, this paper proposes a movie recommendation algorithm based on user profile similarity. First, based on a tag set of movie content features, a user preference profile matrix is constructed by calculating the frequency of user ratings among different movie content tags. Then, user similarity is calculated through this matrix, and a user complex network is modeled to determine the centrality weight of users within the network. Finally, the community weight is obtained through K-core decomposition of the user network. The rating predictions are improved by integrating the centrality weight and community weight of neighboring users. Experimental results show that the proposed algorithm improves prediction accuracy and classification accuracy by 2.72% and 3.17%, respectively. These results validate the effectiveness of complex network modeling based on user profile similarity for enhancing information utilization in recommendation systems.

Motion objects in video frames often exhibit diverse scales over time, which poses a challenge for video anomaly detection. Although traditional generative adversarial networks (GANs) have achieved some success in video anomaly detection tasks, their performance is limited due to the use of a single-scale feature extraction that fails to capture features of objects at different scales. To address this issue, this paper proposes a video anomaly detection method based on a GAN structure that incorporates multi-scale feature fusion and attention mechanisms. Specifically, different-sized convolutional kernels are employed to capture features with varying receptive fields, which are then fused to obtain multi-scale feature representations. Additionally, a coordinate attention mechanism is introduced after the transposed convolutional layers of the generator, allowing adaptive allocation of feature map weights to enhance the model’s perception of crucial features.Experimental results on the public datasets UCSD Ped2 and Avenue demonstrate that the proposed method outperforms existing approaches.

To address the issues of overexposure, color distortion and detail loss in existing enhancement methods when image illumination distribution is uneven, a low-light image enhancement method combining dark region guidance and attention mechanism is proposed. Firstly, the simple linear iterative clustering (SLIC) method is used to generate a dark region guidance map, which guides the network to enhance the underexposed regions of the image while ensuring that the normally exposed regions are not overexposed. Secondly, a channel attention module is designed to improve the extraction of color information, effectively restoring the image color while maintaining natural color fidelity. Subsequently, a global context module is established to enhance the network’s global perception capability, enriching image details. Finally, an enhancement network is designed to fuse the input features with the output features of the dark area attention network,achieving contrast re-enhancement. Multiple comparative experiments are conducted on six public datasets to compare the performance from both subjective and objective aspects. It is shown that the proposed method effectively solves the problems of color distortion, detail loss and uneven exposure in low-light images, delivering superior visual enhancement effect and generalizability.

In this paper, we first investigate the signal and interference characteristics of the windowed universal filtered multi-carrier (UFMC) system in presence of carrier frequency offset (CFO), and then derive the expression of subcarrier signal-to-interferenceplus-noise ratio (SINR). Based on the relationship between symbol error rate (SER) and SINR, an optimization model for window design is formulated to maximize the geometric mean of subcarrier SINR. The successive convex approximation (SCA) algorithm is then employed to solve the optimization problem and determine the optimal window. Finally, SER simulations with various window functions demonstrate that the windowed operations effectively reduce the UFMC interference caused by CFO. Moreover, compared with traditional window functions, the proposed optimal window function significantly improves the system SER, and behaves reliably across a wide signal to noise ratio range.

The soft-clustered capacitated arc routing problem (SoftCluCARP) is an extension of the classical capacitated arc routing problem. Due to its NP-hard nature, solving it is computationally challenging. In this work, we propose an effective hybrid memetic algorithm (HMA) to solve SoftCluCARP. HMA integrates three distinct algorithm modules: a group matching-based crossover to produce promising offspring solutions, a two-stage variable neighborhood search to perform local optimization, and a quality-and-distance population updating to maintain a high-quality population. Experimental results show that HMA is highly competitive compared to the existing exact algorithm in terms of both solution quality and computation time.

Existing graph neural network (GNN)-based question answering (QA) methods using pre-trained language models and knowledge graphs mainly focus on building knowledge graph subgraphs and reasoning processes. However, such methods ignore the semantic differences between question context and knowledge graphs, limiting their ability to deeply mine text representations. Moreover, they fail to comprehensively consider the varying contributions of these two representations to answer prediction. To address these challenges, this paper proposes an expert relearning reasoning QA method based on knowledge graphs and a gating mechanism. This method splices and fuses the question context representation with the inferred knowledge graph representation, and randomly assigns the fused representation vector to the expert network to relearn the entity semantic features associated with the question context and knowledge graph. By mining deeper hidden knowledge and incorporating the gating mechanism, the model accurately scores the question context and the inferred knowledge graph representation, dynamically adjusting their contribution to the answer prediction, and improving prediction accuracy. The proposed method was tested on the CommonsenseQA dataset and OpenBookQA dataset, achieving accuracy improvements of 2.08% and 1.23% over the QA-GNN method, respectively.

To mitigate the significant loss of important geographic entity structural features in mesh simplification algorithms based on quadratic error functions, this paper presents a mesh simplification method that integrates multi-source point cloud feature information. Firstly, laser point clouds and image dense matching point clouds are fused to enhance the quality of the mesh model. Subsequently, incorporating attributes such as color, elevation and curvature, a region-growing algorithm based on super-voxels is applied to segment the fused point cloud and extract feature information. Finally, the quadratic error matrix is updated using the extracted point cloud feature information to achieve high-precision mesh simplification. Utilizing a three-dimensional mesh constructed from the fused point cloud as experimental data, the proposed algorithm is evaluated and compared with QEM, QEF, and Low-poly algorithms. Experimental results indicate that the proposed method improves simplification accuracy by an average of 39.49%.

Traditional copyright management methods rely on centralized servers for storage and verification, which leads to a series of problems, such as difficulty in detecting infringement, complexity in the copyright validation and authorization process, and the absence of an effective similarity retrieval mechanism. These issues make it difficult to provide copyright owners with credible evidence of their copyright. To address these challenges, this paper proposes an image digital copyright protection system based on Hyperledger Fabric blockchain network. The system integrates the scale-invariant feature transform (SIFT) similarity detection algorithm, discrete cosine transform (DCT) zero-watermarking algorithm, chaotic mapping image encryption algorithm, and inter planetary file system (IPFS) for distributed storage. System function tests and performance analyses demonstrate the system’s capability to achieve unique similarity detection, reliable copyright ownership proofs, decentralized encrypted storage, and seamless copyright transfer. These features collectively provide a transparent, safe and open platform for digital image copyright trading, ensuring effective copyright protection for copyright owners.

Entity-relation joint extraction, as a core part of knowledge graph construction, aims to extract entity-relation triples from unstructured text. Current joint extraction methods often struggle with decoding inefficiencies, resulting in weak interaction modeling between entities and relations, insufficient context understanding, and redundant information. To address these limitations, we propose a model based on parallel decoding and clustering for entity-relation joint extraction. First, the bidirectional encoder representations from transformers (BERT) model is used for text encoding to obtain character vectors rich in semantic information. Next, a non-autoregressive parallel decoder is employed to enhance interactions between entities and relations. To further optimize decoding results, hierarchical agglomerative clustering is combined with a majority voting mechanism, improving contextual information capture and reducing redundancy. Finally, a high-quality set of triples is generated to construct a curriculum knowledge graph. To evaluate the performance of the proposed method, experiments are conducted on the public datasets NYT and WebNLG, as well as a self-constructed C language dataset. The results show that this method outperforms other models in terms of precision and F1 score.

Infrared camera images of leopards in natural environments pose significant challenges for individual recognition due to issues such as high fusion between individuals and their surroundings, as well as high inter-class similarity. To address these challenges, an improved EfficientNet model is proposed, incorporating self-calibrating convolution and bilevel routing attention. The self-calibrating convolution adaptively builds remote space and inter-channel dependencies around each spatial location. The ability to recognize detailed features is enhanced by explicitly combining richer contextual information. This effectively mitigates the recognition challenges posed by high inter-class similarity. Meanwhile, the bilevel routing attention combines the top-down global attention strategy and the bottom-up local attention strategy to solve the problem of high integration between individuals and their environment. Experiment results show that the accuracy of the proposed model reaches 95.56% in the task of leopard individual recognition, which is significantly higher than the original EfficientNet. These findings validate the effectiveness and superiority of the proposed model in dealing with leopard individual recognition task.