Search results

Vision transformers (ViT) detectors excel in processing natural images. However, when processing remote sensing images (RSIs), ViT methods generally exhibit inferior accuracy compared to approaches based on convolutional neural networks (CNNs). Recently, researchers have proposed various structural optimization strategies to enhance the performance of ViT detectors, but the progress has been insignificant. We contend that the frequent scarcity of RSI samples is the primary cause of this problem, and model modifications alone cannot solve it.

To address this, we introduce a faster RCNN-based approach, termed QAGA-Net, which significantly enhances the performance of ViT detectors in RSI recognition. Initially, we propose a novel quantitative augmentation learning (QAL) strategy to address the sparse data distribution in RSIs. This strategy is integrated as the QAL module, a plug-and-play component active exclusively during the model’s training phase. Subsequently, we enhanced the feature pyramid network (FPN) by introducing two efficient modules: a global attention (GA) module to model long-range feature dependencies and enhance multi-scale information fusion, and an efficient pooling (EP) module to optimize the model’s capability to understand both high and low frequency information. Importantly, QAGA-Net has a compact model size and achieves a balance between computational efficiency and accuracy.

We verified the performance of QAGA-Net by using two different efficient ViT models as the detector’s backbone. Extensive experiments on the NWPU-10 and DIOR20 datasets demonstrate that QAGA-Net achieves superior accuracy compared to 23 other ViT or CNN methods in the literature. Specifically, QAGA-Net shows an increase in mAP by 2.1% or 2.6% on the challenging DIOR20 dataset when compared to the top-ranked CNN or ViT detectors, respectively.

This paper highlights the impact of sparse data distribution on ViT detection performance. To address this, we introduce a fundamentally data-driven approach: the QAL module. Additionally, we introduced two efficient modules to enhance the performance of FPN. More importantly, our strategy has the potential to collaborate with other ViT detectors, as the proposed method does not require any structural modifications to the ViT backbone.

This paper aims to make a comprehensive evaluation of the progress and achievements made by China and Japan in literature collection, publication and academic research of South Manchuria Railways Co (SMR) after the Second World War and points out the existing and urgent problems.

This paper adopts the methods of literature research, questionnaire survey, comparative analysis and expert consultation.

After the Second World War, both China and Japan made great achievements in the collection, publication and research of SMR data, as well as in the academic research of SMR. However, as a new research field, there are still some problems in deepening the excavation and utilization of SMR data, expanding the research fields, improving the research methods, etc.

SMR itself was the product of a war of aggression. Although it died with the defeat of the war of aggression, remaining SMR data also has very precious value, it records the history of Japanese aggression against China, is the biggest database for the study of the problems in China and northeast Aisa at that time. In addition to Japan and China, the USA and the Soviet Union also kept part of the SMR data, so that to involve them in the research of SMR is also the purpose of this paper.

This paper aims to investigate the performance of Jiangsu Province industrial technology innovation strategy alliance.

Through a preliminary investigation of 30 Jiangsu industrial technology innovation strategic alliances, this paper analyzed the status and extracted 18 alliances to conduct an in-depth investigation. By grey evaluation method based on center-point triangular whitenization weight function, the paper classified and analyzed alliances.

The results show that university or research institutions-oriented alliance perform better, but the government/enterprise-oriented alliance perform diverse, and majority is rated “general”.

The paper succeeds in clustering analysis to Jiangsu Province industrial technology innovation strategy alliance with insufficient data. And according to the result of clustering, it analyzes the causes, which provide value information for the sustainable development of Jiangsu Province industrial technology innovation strategy alliance.

The paper aims to examine the relationships among marathon spectator motivation, satisfaction and behavioral intention in China.

Date were collected from three Chinese marathon races. The paper presents the conceptual model of the study and adopted a mixed method to identify the motivations of marathon spectators; structural equation modeling (SEM) was used to test the measurement and substantive models; regression analysis was used to detect the specific effects of the various spectator motivations on spectator satisfaction and behavioral intention.

The paper provides empirical evidences about how marathon spectator motivation affects the behavioral intention. The research results from a path analysis of the spectator motivation–satisfaction–behavioral intention model show that: spectator motivation has significant positive effects on behavioral intention, the indirect path dominates the total effect of motivation on behavioral intention via satisfaction as a mediating variable; satisfaction is positively associated with behavioral intention, and it plays mediate role between spectator motivation and behavioral intention. The result of regression analysis indicates that various spectator motivations have different influences on the satisfaction and behavioral intention, motives of excitement and diversion have significant effect on satisfaction and behavioral intention, but socialization motive is not significantly associated with the satisfaction and behavioral intention.

This paper first fulfills an identified need to explore mass sports event spectator motivations and its influential mechanism on behavioral intention among Chinese marathon races by survey. It contributes greatly to the literature of mass sports spectator motivation and consumption theory, provides important managerial implications for Chinese sports organizations as to how to improve spectator orientation degree.

Grounded in upper echelons theory (UET) and conflict theory, the purpose of this study is to analyzes the innovative behavior of family firms from the perspective of the cognitive differences between successors and senior managers.

This research employed a sample listed family firms in China. The obtained results were subjected to hierarchical regression analysis, complemented by rigorous model robustness testing through propensity score matching and regression with substitution variables.

Successors engender task conflicts with family members in the top management team (TMT) due to cognitive differences, thereby stimulating corporate innovation. Conversely, successors engender relationship conflicts with non-family members in the TMT, impeding innovation. Furthermore, the performance expectations and the gender of the successor CEO also influence the relationship between cognitive differences and innovation between the successor and the TMT.

This study's originality and value lie in its innovative application of UET and conflict perspectives to dissect the intricate layers of intergenerational cognitive differences and their impact on the innovative behavior of family firms. It augments our comprehension of how the internal dynamics within family firms shape strategic innovation decisions.

Despite the importance of demand forecasting in retail industry, its influence on supply chain agility has not been sufficiently examined. From a total information technology (IT) capability perspective, the purpose of this paper is to examine the antecedent of supply chain agility through retail demand forecasting.

Combining the literature reviews, the quantitative method of algorithm analysis was targeted at, and the firm data were processed on MATLAB.

This paper summarizes IT dimensions of demand forecasting in retail industry and distinguishes the relationship of supply chain agility and demand forecasting from an IT capability view.

Managers can derive a better understanding and measurement of operating activities that appropriately balance among supply chain agility, IT capability and demand forecast practice. Demand forecasting should be integrated into the firm operations to determine the agility level of supply chain in marketplace.

This paper constructs new theoretical grounds for research into the relationship of demand forecasting-supply chain agility and provides an empirical assessment of the essential components for the means to prioritize IT-supply chain.

This study aims to explore the value of pocket parks and associated factors, as well as the differences between pocket parks and other green spaces, and finally to identify the role of pocket parks in cities.

This study adopts a systematic literature review method to identify 85 articles from the Web of Science and other sources. The Nvivo12plus software was employed to qualitatively analyse the value of pocket parks, and the key associated factors were visualised via the NVivo word cloud function.

As a result, pocket parks reflected high social, environmental ecological, cultural aesthetic values but low economic value. Compared to other large-scale green spaces, the particularity of pocket parks lies in their ambiguity and temporality of space and high reliance on the surrounding environment. Small and scattered pocket parks in the city are like “convenience stores” and “green magnets”, and connectivity should be strengthened to promote urban regeneration.

Compared with previous studies that mostly focused on large urban green spaces, the originality of this study is that it elaborates on the importance and values of pocket parks and proposes the particularity and roles of pocket parks in cities, which enriches the theoretical knowledge of pocket parks. In terms of practical applications, the proposed diverse values and associated factors contribute to guiding the planning, management and operation of pocket parks. Moreover, the roles of pocket parks help enhance park connectivity and establish a park network system.

The purpose of this study is to explore the relationship between big data and supply chain platform in China’s retail industry. With the emergence of big data resources and technologies, the business pattern of new retail advocates the combination of online and offline channels. Supply chain platform plays a key role in the implementation of retail activities, which has gradually become a research hotspot in the cross field of operations management and information system.

Through the method of literature review and case study, this study empirically explores how big data shapes supply chain platform to support new forms of online retail by grounded theory.

The model framework is validated by reliability test and coding method to process survey materials. The results identify the overall antecedents of supply chain platform and reveal positive effects between big data and new retail. The findings help firm managers build a big data-driven supply chain to support new retail.

There are insufficient studies on theoretical frameworks and interaction relationships among big data, supply chain platform and new retail.

Live-streaming platforms emphasize dynamic social interaction and fan engagement. Users integrate into the live-streaming community through continuous social learning activities, such as sending bullet comments, reviewing comments and interacting with celebrities. However, comprehensive research on the transactional intricacies of live-streaming e-commerce from the perspectives of community and learning is still lacking.

Focusing on the behavior characteristics of the reference group represented by online celebrities and fans in the live-streaming environment, this study utilized social learning as the theoretical basis to examine how reference groups affect consumer purchase intention through a series of intermediary effects. An empirical investigation and machine learning algorithms were utilized to explore and verify the hypothesized model.

The results show that: (1) reference groups’ behavior positively stimulated social presence and enhanced consumer purchase intention through the chain-mediating effect of social presence and trust in online celebrities; (2) celebrity characteristics (professionalism, attractiveness and interactivity) positively impacted consumer trust; (3) in addition, machine learning algorithms substantiated that reference groups’ behavior, social presence, trust and celebrity characteristics had a remarkably robust predictive effect on purchase intention.

These findings hold theoretical implications for understanding how the social community affects consumers’ purchase intention in the live-streaming context and practical significance for marketing strategies toward live-streaming e-commerce.