Concurrency and Computation-Practice & Experience最新文献

The ability to generate facial expressions is essential for humanoid social robots to engage in natural, human-like interactions. This capability significantly enhances the fluidity of human-robot communication and the precision of emotional expression. However, current methods rely heavily on pre-programmed behavioral patterns, which are manually implemented at considerable cost in both time and human labor. To enable humanoid robots to autonomously acquire generalized expressive capabilities, they must learn human-like expressions via self-supervised training. To address this challenge, we present a highly biomimetic robotic face equipped with physically actuated electronic facial units, alongside an end-to-end learning framework that integrates Kolmogorov-Arnold Networks (KAN) with attention mechanisms. In contrast to previous approaches, we have also developed an automated data collection system guided by expert-designed facial motion primitives, enabling the construction of a high-quality dataset. Notably, to the best of our knowledge, this constitutes the first facial expression dataset specifically designed for humanoid social robots. Extensive evaluations demonstrate that our method enables accurate and diverse facial mimicry across a range of test subjects.

This paper proposes an intelligent and adaptive security framework for internet of things (IoT) environments by integrating edge Honeypots, AI-driven intrusion detection systems (IDS), and software-defined networking (SDN). The system is designed to enhance real-time threat detection, reduce false positives, and dynamically mitigate attacks. Edge Honeypots are deployed at the network perimeter to attract and analyze malicious traffic, which is then used to train AI-based IDS models. The IDS employ a hybrid detection mechanism combining signature-based and anomaly-based techniques. SDN facilitates centralized traffic control and dynamic rule updates, enabling rapid responses to new attack vectors. The framework is implemented and evaluated in a simulated SDN-IoT environment using Mininet, with several machine learning models benchmarked. The Decision Tree model achieves the highest detection accuracy (97%) for IoT threats. Experimental results demonstrate improved detection performance, reduced false positives, and enhanced adaptability through a continuous learning loop between the IDS and honeypots.

Small object detection, a critical technique for recognizing and localizing diminutive targets in visual data, plays a vital role in applications ranging from remote sensing and unmanned aerial vehicle (UAV) vision to autonomous driving. Current methodologies, however, face substantial challenges, including detection accuracy limitations, low-resolution image processing difficulties, background noise interference, and target occlusion issues. To address these challenges, we propose FMF-DETR, an innovative small object detection framework featuring frequency-domain feature optimization through three key components: (1) High-Low Frequency Fusion Model (HLFM), (2) Focused Diffusion Feature Pyramid Network (FDFPN), and (3) BiPathNet (BPNet). Specifically, the HLFM module enhances multiscale feature representation by emphasizing high-frequency details while suppressing low-frequency background noise. The FDFPN architecture improves detection performance in complex scenarios through multiscale feature fusion and saliency-aware diffusion. BPNet introduces a dual-path feature extraction mechanism that simultaneously enhances feature discriminability and reduces computational overhead. Through the synergistic integration of these components, the proposed framework enhances both detection accuracy and operational efficiency. Comprehensive evaluations on the VisDrone dataset demonstrate FMF-DETR's superior performance, achieving a 2.2% accuracy improvement while reducing model parameters by 13.03M and computational complexity by 97.2G FLOPs compared to baseline methods. These results validate both the effectiveness and efficiency of our proposed framework.

Accurate fuel consumption prediction is critical for effective supply chain management and decision-making in industrial engineering systems. In price commitment scenarios, the discrepancies between market prices and transaction prices necessitate the development of scenario-specific prediction methods. However, such scenarios present several challenges. Market price fluctuations can induce variations in consumption patterns, while transaction price dynamics often differ from those in traditional markets, rendering conventional prediction models inadequate. Additionally, the cold-start characteristic of business results in limited historical data, and traditional methods relying on random sampling strategies often neglect sample heterogeneity, leading to biased weight allocation. To address these issues, we propose an Adaptive Sampling LightGBM framework (ASL), which integrates an adaptive sampling strategy and post-sampling bagging to enhance model stability with limited samples. By constructing a scenario-oriented risk function, the method incorporates market price volatility intensity and the divergence between commitment and conventional scenarios as dynamic weighting factors, thereby quantifying the time-varying impact of price elasticity on consumption. Experimental results using data from two fuel stations in North China demonstrate that our method outperforms benchmark prediction methods.

In the current panorama of digital healthcare evolution and the proliferation of electronic health records (EHRs), healthcare systems face many challenges. These challenges encompass data administration and security, data exchange, and ensuring the confidentiality of medical data. Within this context, blockchain technology emerges as a promising way to address the numerous challenges associated with EHRs. Ensuring the security and confidentiality of EHRs remains a pivotal concern encompassing healthcare service recipients and providers. The compromise of a healthcare system leads to the exposure of intricately private health information. Typically stored in centralized databases, this information repository introduces susceptibilities that consequently fuel instances of cyber intrusion. To ensure the safeguarding of medical data, this research paper strategically adopts the Hyperledger Fabric (HLF) blockchain platform, which incorporates attribute-based access control mechanisms to thwart any malicious attempt at unauthorized access to sensitive information. These apprehensions regarding security have been effectively addressed by leveraging robust cryptographic protocols such as the AES-256 algorithm. This algorithm encrypts messages, transmitting the encrypted data across the network, thereby restricting visibility solely to intended recipients and providing a secure data exchange. The effectiveness of our proposed solution is rigorously evaluated using the Hyperledger Caliper tool. The evaluation encompasses pivotal performance metrics, average latency, throughput, success rate, resource consumption, and traffic.

Visible-infrared person re-identification (VI-ReID) is a cross-modal retrieval task characterized by significant challenges, with the objective of precisely identifying and matching pedestrian instances across different spectral modalities, namely visible-light and infrared imagery. The primary difficulties stem from substantial inter-modal discrepancies and intra-modal feature variations, which complicate effective cross-modal matching. While existing approaches generally focus on embedding heterogeneous modal data into a unified feature space to extract shared representations, they often overlook the discriminative identity information embedded within modality-specific features. To overcome this inherent limitation, we propose a novel pinwheel-guided dynamic representation network (PDRNet), designed to mine and enhance the directional structural cues and scale-sensitive discriminative features inherent in modality-specific representations. Specifically, we integrate direction-aware pinwheel convolution (PConv) into a two-stream architecture to strengthen local structural representation and guide the learning of shared semantic features, thereby improving both the discriminability and structural modeling of modality-specific information. Furthermore, to accommodate scale disparities across modalities and individuals, we incorporate a scale-based dynamic loss (SD loss), which adaptively adjusts the loss weights related to scale and positional information. This mechanism mitigates the error amplification often observed in small-scale samples and enhances both the discriminative power and robustness of cross-modal matching across varying scales. We perform extensive experiments on multiple well-established public benchmarks. The results consistently show that the proposed PDRNet achieves superior performance compared to existing methods in both recognition accuracy and cross-modal matching effectiveness.

Multimodal sentiment analysis (MSA) has recently encountered two major challenges: non-textual modalities are often affected by noise, and sentiment intensity differences are difficult to capture. To address these issues, we propose a Sentiment Intensity Contrastive Text-Enhanced Fusion Network (SICTEF Net), which achieves deep collaboration among text, audio, and visual modalities through three key mechanisms. First, a grouped-channel-attention based Feature Enhancement Module (EMA) is designed to mitigate modality-specific noise and emphasize emotion-sensitive cues by combining spatial–channel interaction mapping with dual-branch attention fusion. Second, a text-centered cross-modal fusion mechanism is introduced, where bidirectional multi-head self-attention and a residual-enhanced encoder jointly enable complementary mappings between text and non-text modalities, thereby producing intermediate representations that preserve semantic primacy while incorporating fine-grained complementary information. Third, a sentiment-intensity weighted contrastive learning strategy dynamically assigns weights to positive and negative sample pairs according to their sentiment intensity differences, allowing the model to more precisely distinguish samples with varying degrees of similarity in the embedding space. Experimental evaluation on the CMU-MOSI and CMU-MOSEI datasets demonstrates that SICTEF Net consistently outperforms state-of-the-art baselines in binary accuracy, F1 score, seven-class accuracy, mean absolute error (MAE), and Pearson correlation. Comprehensive ablation studies further confirm the complementary benefits of EMA, the text-enhanced Transformer, and sentiment-intensity contrastive learning. These results indicate that combining text-driven deep interaction, non-text modality enhancement via channel attention, and contrastive learning can improve the accuracy and robustness of multimodal sentiment analysis.

Mobile application recommendation has emerged as a pivotal domain within the realm of personalized recommendation systems. Traditional mobile application sequence recommendation approaches are predominantly dedicated to the pursuit of sophisticated sequence encoders to achieve more precise representations. However, existing sequence recommendation methods primarily consider the sequential order of historical App interactions, overlooking the time intervals between applications. This oversight hinders the model's capability to fully unearth the temporal correlations in user behavior, consequently limiting the accuracy and personalization of mobile application recommendations. Moreover, the interactions between users and mobile applications are typically sparse, which weakens the model's generalization capabilities. To address these issues, we propose a novel method for mobile application sequence recommendation, incorporating time interval-aware attention and contrastive learning (called Ti-CoRe). Specifically, this approach introduces a novel sequence augmentation strategy based on similarity replacement within a contrastive learning framework. By considering textual similarities between applications, this method selectively replaces applications that possess lower similarity scores to generate augmented sequences, increasing the diversity of the sample space and mitigating data sparsity. Furthermore, integrating a time interval-aware mechanism into the BERT4Rec model, the paper presents a new T-BERT encoder. It precisely assesses the influence of fluctuating time intervals on the prediction of the subsequent mobile application, thereby ensuring a more nuanced app representation. Experiments conducted on the 360APP real dataset demonstrate that Ti-CoRe consistently outperforms various baseline models in terms of NDCG and HR metrics.