Salient object detection (SOD) in optical remote sensing images (ORSIs) faces numerous challenges, including significant variations in target scales and low contrast between targets and the background. Existing methods based on convolutional neural networks (CNNs) and vision transformers (ViTs) attempt to exploit both local details and global dependencies; however, their heterogeneous feature representations remain difficult to integrate effectively, limiting overall performance. To overcome these limitations, we propose an Adaptive State-Space Context Network (ASCNet), which adapts visual state-space modeling to ORSI-SOD within a unified encoder–decoder architecture. ASCNet employs a visual state-space encoder to extract hierarchical features and introduces a Multi-Level Context Module (MLCM) to align adjacent-level features and perform topology-aware regional aggregation, producing context-enhanced skip representations. For decoding, we design an Adaptive Patchwise Visual State Space (APVSS) block composed of two components. The Granularity-aware Propagation Module (GPM) inserts learnable global tokens into local patch sequences and modulates patch responses before state-space propagation. The Dynamic Adaptive Granularity Scan (DAGS) constructs scan sequences through resolution-aware block partitioning and content-adaptive directional weighting, enabling low-resolution features to retain global layout information while allowing high-resolution features to focus on local geometric structures. Extensive experiments on ORSSD, EORSSD, and ORSI-4199 demonstrate that ASCNet consistently outperforms recent SOD and ORSI-SOD methods, validating the effectiveness of multi-level context aggregation and adaptive patchwise state-space decoding.
In order to enhance the service life of TC4 alloy under harsh high-temperature conditions, CrNiMnTiAl-yTi3SiC2 (y = 5, 7.5, 10 wt%; TS5, TS7.5 and TS10) composite coatings were fabricated on TC4 alloy via laser cladding additive manufacturing. The microstructural evolution, tribological properties, and high-temperature oxidation resistance were investigated. The results show that the TS7.5 exhibits the best high-temperature wear resistance, with a 62% reduction in wear compared to the substrate, corresponding to an approximately four-fold improvement in wear resistance at 600 degrees C. This is attributed to the internal dispersion strengthening, together with the synergistic effect of a uniformly distributed dislocation network structure and a high proportion of high-angle grain boundaries, which effectively hinder dislocation motion, alleviate local stress concentration, and significantly enhance the coating strength and wear resistance. Besides, the TS10 demonstrates superior oxidation resistance at 800 degrees C, achieving a 59% reduction in oxidation weight gain relative to the substrate. This results from the higher Ti3SiC2 content provides sufficient SiO2, which integrates with Cr2O3, TiO2, Al2O3 and other oxides to construct dense composite oxide scale that effectively suppresses oxidation reactions. By investigating the underlying mechanisms of tribological properties and oxidation resistance, this work not only extends the service life of TC4 alloy but also provides new theoretical support for enhancing coating performance.
Abstract Alternative dispute resolution (ADR) mechanisms are considered to offer several advantages over litigation and arbitration, including greater flexibility, lower cost, and better preservation of cooperative relationships between disputing parties. However, they have not been widely adopted in China’s construction industry. This study aims to identify the barriers to the implementation of ADR mechanisms for dispute resolution in the Chinese construction industry. First, 15 barrier factors were identified through a literature review and expert interviews. Then, the integrated Decision Making Trial and Evaluation Laboratory-Interpretive Structural Modeling-Cross Impact Matrix Multiplication Applied to Classification (DEMATEL-ISM-MICMAC) approach was employed to analyze the interrelationships among these factors and to develop a multilevel hierarchical structural model of the barriers to implementing ADR mechanisms for dispute resolution in China’s construction industry, thereby identifying the most fundamental barriers. The results were further validated based on the driving power and dependence of the barrier factors. The findings indicate that the lack of supporting laws and regulations is the most fundamental barrier to the implementation of ADR mechanisms in China’s construction industry. The absence of explicit ADR provisions in contracts and the low willingness to use ADR mechanisms are identified as direct barriers. The results are of great significance for enriching the body of knowledge on ADR mechanisms and promoting their global development. They also provide useful references for the promotion of ADR mechanisms in China and in other countries and regions where ADR is still at an early stage of development.
Blind super-resolution (SR) of remote sensing images is highly challenging due to the complex and spatially variant degradations caused by diverse imaging conditions, atmospheric effects, and sensor characteristics. Most existing methods assume uniform degradation across the entire image or rely on global degradation representations, which fail to capture the intrinsic spatial heterogeneity in real-world remote sensing scenarios. To address this issue, a novel spatial degradation field modeling framework is proposed, in which spatially varying degradations are explicitly characterized at the patch level. Specifically, a spatially adaptive degradation estimation network (SADEN) is designed to estimate patch-wise degradation representations, which are further organized into a spatial degradation field to describe the degradation distribution across the image. Based on this representation, an adaptive degradation adjustment module (ADAM) is introduced to dynamically modulate feature responses conditioned on the estimated degradation. The module incorporates both spatial modulation (SM) and channel modulation (CM) mechanisms, enabling the network to adaptively refine spatial structures and channel-wise feature responses under complex degradation patterns. Extensive experiments across multiple benchmark remote sensing datasets demonstrate that the proposed method effectively captures spatially varying degradation patterns and consistently outperforms state-of-the-art blind super-resolution approaches in both quantitative performance and visual quality. Moreover, its effectiveness is further validated through evaluations on real-world scenarios as well as downstream semantic segmentation tasks.
Unfair negative reviews can harm businesses by damaging brand image and reducing customer trust. This research examines how tourism and hospitality businesses should respond to different types of unfair reviews through three scenario-based experiments. The findings reveal a congruity effect: rational responses are more effective in improving third-party consumers’ purchase intentions for low-arousal unfair reviews, while emotional responses are more effective for high-arousal ones. The results further show that social distance mediates this effect. In addition, in high-arousal cases, weak victimization responses are more suitable than strong reprimands. These findings enrich theories related to Communication Accommodation Theory and contribute to the literature on unfair negative reviews. They also provide practical guidance for tourism and hospitality businesses on how to effectively handle online consumer unfair reviews.
