Search Results (3)

Aiming to address the fundamental limitation of fixed activation functions that constrain network expressiveness in existing deep deblurring models, in this pioneering study, we introduced Kolmogorov–Arnold Networks (KANs) into the field of full-color/RGB image deblurring, proposing the Kolmogorov–Arnold and Transformer Network (KT-Deblur) framework based on dynamically learnable activation functions. This framework overcomes the constraints of traditional networks’ fixed nonlinear transformations by employing adaptive activation regulation for different blur types through KANs’ differentiable basis functions. Integrated with a U-Net architecture within a generative adversarial network framework, it significantly enhances detail restoration capabilities in complex scenarios. The innovatively designed Unified Attention Feature Extraction (UAFE) module combines neighborhood self-attention with linear self-attention mechanisms, achieving synergistic optimization of noise suppression and detail enhancement through adaptive feature space weighting. Supported by the Fast Spatial Feature Module (FSFM), it effectively improves the model’s ability to handle complex blur patterns. Our experimental results demonstrate that the proposed method outperforms existing algorithms in terms of peak signal-to-noise ratio (PSNR) and structural similarity (SSIM) metrics across multiple standard datasets, achieving an average PSNR of 41.25 dB on the RealBlur-R dataset, surpassing the latest state-of-the-art (SOTA) algorithms. This model exhibits strong robustness, providing a new paradigm for image-deblurring network design. Full article

Urban Air Mobility (UAM) is an emerging industry marked by technological advancements, new operational contexts, and regulatory frameworks. This article examines how to improve safety management in UAM operations by adopting a just culture approach from a system of systems perspective. Acknowledging the critical role of front-line workers, especially in the early stage with piloted vehicles, the ecosystem-level approach comprehends multiple providers, operators, and services. Employing an enterprise architecture methodology, we address the challenge of fostering a learning-oriented environment amidst diverse organizational perspectives and stakeholders’ interests. This study identifies key capabilities, functions, and resource exchanges within and across organizations by strategically leveraging architectural views and systemic visualizations. A unified safety committee is discussed and recommended to facilitate consensus among stakeholders, including regulatory bodies, thus paving the way for industry-wide improvements. Findings contribute to evolving safety protocols in UAM operations and serve as a blueprint for integrating cutting-edge methodologies to drive systemic enhancements. Full article

System-of-systems (SoS) architecture is crucial in managing complex and interconnected systems. However, the description and modeling of SoS architecture pose significant challenges and require a structured and organized approach. In this study, a metamodel for SoS architecture that considers both structural and behavioral perspectives is defined. The metamodel is then mapped to ontologies that are enhanced with a flow-based extension to characterize architecture views. On this basis, an SoS capability ontology (SoSCO) and an SoS operational ontology (SoSOO) are built with factors, relationships, and flows. A four-step architecture process for developing capabilities and a five-step architecture process for operational modeling are provided based on the ontologies. The proposed approach is applied in a search and rescue case study, demonstrating its ability to improve operability in the early design stage. The process is implemented using the Unified Architecture Framework (UAF) so that various stakeholders and engineers can better understand and develop an SoS. Full article