安徽财经大学

Text-based person re-identification (TBPReID) aims to retrieve person images from a gallery using natural language descriptions. Despite recent progress, TBPReID remains challenging because image–text pairs in real-world datasets often suffer from noisy correspondence (NC), including ambiguous descriptions, attribute-level inconsistencies, and mismatched image-caption pairs. Such unreliable supervision may force cross-modal models to align semantically inconsistent samples and degrade fine-grained retrieval performance. To address this problem, we propose an Uncertainty-Aware Alignment (UAA) framework for robust TBPReID under NC. The key idea is to model image–text correspondence as a continuous reliability-estimation problem and propagate the estimated reliability into fine-grained alignment. Specifically, we introduce an Uncertainty-Aware Consensus (UAC) mechanism, where consensus denotes the reliability of cross-modal matching inferred from in-batch similarity evidence. UAC constructs Dirichlet-based evidence distributions to estimate epistemic uncertainty and adaptively down-weight unreliable correspondences. We further design a Fine-Grained Attribute Masking (FGAM) module with an uncertainty-weighted ranking loss, which uses uncertainty-derived confidence to suppress ambiguous attributes and strengthen trustworthy identity-related cues. Extensive experiments on CUHK-PEDES, ICFG-PEDES, and RSTPReid demonstrate the effectiveness and robustness of UAA. Under the 50% noise setting, UAA achieves 71.58% R@1 on CUHK-PEDES, yielding a 4.01% improvement over the baseline. Our code is available at https://github.com/pengchengL-hub/UAA.

Customer referral programs are a cornerstone of growth strategies, predicated on the assumption that loyal customers are universally effective advocates. However, research on the boundary conditions under which loyalty translates into referral behavior remains limited. In collaboration with a major Chinese e-commerce firm, we conducted a field experiment using Cox proportional hazards models. Our findings reveal that social value-framing content acts as a critical boundary condition: the interaction between customer loyalty and social value-framing content negatively impacts referral intention. Specifically, the referral advantage associated with loyal users is attenuated under social value-framing content, a phenomenon we term “loyal but not virtuous”. Supplemental experiments reveal that this effect is mediated by self-enhancement motivation. For high-loyalty users, their established identity as brand advocates creates a potential image threat when sharing social-value content, which suppresses their drive for self-enhancement. This study enriches the literature on customer loyalty and referral programs and provides practical recommendations for firms on aligning loyalty with content strategies.

Climate change has become one of the most serious threats to environmental sustainability, economic stability, and long-term social welfare, making climate resilience an essential policy objective. In this context, reducing the severity of climate change requires a better understanding of the structural drivers of carbon emissions and the institutional conditions that support effective mitigation. This study investigates the effects of environmental policy stringency(EPS), exports, political risk, economic growth, energy efficiency, and policy uncertainty on energy-related CO2 emissions(ERE) in the United States from 1985 to 2021. Using annual time-series data, the study applies Fully Modified Ordinary Least Squares and Dynamic Ordinary Least Squares to estimate the long-run relationships among the variables. The results show that stronger EPS and improved energy efficiency significantly reduce ERE, indicating that regulatory commitment and technological improvement are central to climate resilience. The findings also reveal that greater political stability supports emissions reduction by enabling governments to implement credible, consistent, and long-term climate strategies. At the same time, policy uncertainty and growth-related pressures remain important factors shaping emissions dynamics, while trade-related effects continue to influence the carbon intensity of economic activity. By identifying the conditions under which emissions can be reduced more effectively, the study demonstrates how stable institutions, efficient energy use, and EPS policies can help abate climate severity by limiting the drivers of global warming. Overall, the paper argues that strengthening regulatory certainty, promoting clean energy investment, and improving institutional stability are critical for enhancing climate resilience and accelerating the transition to a lower-carbon economy in the United States.

Drawing on Conservation of Resources theory, this study divides perceived supervisor trust into two dimensions: perceived supervisor dependence and perceived supervisor disclosure, and explores its double-edged sword effect on employee bootleg innovation. The study further investigates the moderating role of future work self-salience from the perspective of the resource balance mechanism.Based on data from 404 valid questionnaires, the results indicate that perceived supervisor dependence is positively associated with employee bootleg innovation, whereas perceived supervisor disclosure is negatively associated with it. Moreover, self-expectations for creativity mediate the relationship between perceived supervisor dependence and bootleg innovation, while role stress mediates the relationship between perceived supervisor disclosure and bootleg innovation. In addition, future work self-salience positively moderates the effect of perceived supervisor dependence on self-expectations for creativity, and negatively moderates the effect of perceived supervisor disclosure on role stress. These findings offer both theoretical insights and practical guidance for managers in effectively managing employee bootleg innovation.

This study investigates the impact of foreign direct investment (FDI) entry on the low-carbon transition of the global supply chain (GSC) in China. We proxy for the global carbon footprint by estimating the carbon emission intensity embedded in firms’ GSCs. Using a difference in differences approach that exploits adjustments to the FDI Catalogue as a quasi-natural experiment, we find that FDI entry reduces the global carbon footprint of domestic firms. We also observe that foreign firms exhibit lower trade-induced carbon emission intensity than domestic firms do, which supports the pollution halo hypothesis. This study contributes to FDI spillovers by exploring the supply chain sharing mechanism linked to​ the variety, quality, and origin of intermediate imports and the technology spillover channel associated with green innovation and green product development. FDI spillovers also occur through vertical connections between industries. Moreover, although the exit of foreign-owned enterprises weakens the spillover effects of FDI entry, their conversion to domestic ownership strengthens these effects.