898 papers
This systematic literature review of 136 articles identifies 14 key research perspectives on the quality and understandability of Unified Modeling Language (UML) notation, reveals a growing trend toward testing and evaluation, and proposes future research directions including integration with low-code platforms and enhanced automated assessment.
This paper presents the Interpretable Edge Intelligence (IEI) framework, which combines an LSTM-Autoencoder with K-Means-compressed KernelSHAP to achieve real-time, causal root cause analysis of QoS anomalies in 5G private networks, demonstrating high accuracy and the critical necessity of local calibration to prevent deployment failures.
This paper proposes the Brain–Behavior Alignment Network with Cross-Modal Contrastive Learning Transformer (BBAN-CMCLT), a novel multimodal deep learning model that integrates fMRI, facial, eye-tracking, and clinical data to achieve state-of-the-art accuracy (98.97%) in detecting Autism Spectrum Disorder by effectively aligning brain and behavioral patterns.
The paper proposes CPBS, a collaborative proxy blind signature scheme that leverages a (t,n)-threshold mechanism to distribute trust among decentralized peer nodes for secure, privacy-preserving, and auditable data collection in resource-constrained Medical IoT networks, significantly reducing computational and communication overhead compared to existing centralized solutions.
This paper proposes CASPA, a novel super-resolution network that combines Content-Aware Global Aggregation and Spatial prior Channel Attention modules to overcome the limitations of existing Vision Transformers, thereby achieving high-fidelity reconstruction of Thangka images with enhanced long-range semantic capture and fine geometric detail preservation.
This paper introduces Difficulty-Aware Sample Allocation (DASA), an architecture-agnostic framework that enhances semantic segmentation performance by dynamically assigning stronger data augmentation to samples based on a multi-factor difficulty score combining prediction ambiguity, training loss, class rarity, and boundary complexity.
This study demonstrates that pretrained visual encoders like CLIP and SigLIP can detect persistent stylistic signals across sketches and paintings better than handcrafted descriptors, yet their fragility against visually similar negatives suggests they are more valuable as tools for analyzing AI's perception of artistic style than as reliable authentication mechanisms.
This paper proposes and evaluates a hardware-assisted zero-knowledge authentication scheme using a USBKEY and the GSVOLE-2DLC protocol to enable secure, privacy-preserving, and lightweight identity verification for resource-constrained IoT terminals.
This paper introduces CQ-Transformer, a hybrid extractive question answering framework that combines query decomposition with ALBERT-based semantic filtering and RoBERTa/XLNet encoders to significantly outperform existing baselines in handling complex, multi-category queries on the SQuAD 2.0 dataset.
This paper proposes a dual-branch late fusion model that integrates ONOS controller telemetry with data-plane flow statistics using the ASEADOS-SDN-IoT benchmark, achieving high detection accuracy while significantly reducing the packet count required for early attack identification compared to data-plane-only baselines.