Search

Shadow detection is significant for scene understanding. Ambiguities in a shadow image, such as shadow-like non-shadow regions and shadow regions with non-shadow patterns, are still very challenging for prevalent CNN-based methods. This work attempts to alleviate this problem from a new perspective of shape semantics, and then proposes a Semantic-aware Transformer (SaT) in a multi-task learning manner. Concretely, we first propose a shadow detection network based on the recent progress of Transformer architecture, allowing us to capture significant global interactions between contexts. Next, we design a multi-task learning framework, combining shadow supervision and semantic supervision to perform a semantic-aware shadow detection. Finally, we introduce a simple yet effective information buffer unit to overcome the gradient signal conflict from multi-task learning. Experimental results on three public benchmark datasets (i.e., ISTD, SBU, and UCF) show that our SaT can effectively detect ambiguous cases and achieve state-of-the-art results. • We propose a shadow detection method using shape semantics. • We design an efficient shadow detection network based on Transformer. • We construct a multi-task learning framework combining shadow and semantic supervision. • An information buffer unit is proposed to coordinate different supervised tasks. [ABSTRACT FROM AUTHOR]

The global imperative to mitigate carbon emissions for sustainable development has spurred extensive research into economic, social, and energy-related factors. However, prior studies present a complex landscape, yielding mixed conclusions regarding the influence of geopolitical risk, natural resource rents, corrupt governance, and energy intensity. To untangle this ambiguity, we construct a research model grounded in the Environmental Kuznets Curve, employing panel data from 38 countries spanning 2002 to 2020. Employing panel quantile regression models, we directly assess the impact of identified factors. Our findings affirm the alignment between economic growth and carbon emissions, supporting the Environmental Kuznets Curve hypothesis. Notably, increased geopolitical risk and energy intensity correlate with heightened carbon emissions over time, while corruption governance and natural resource rents exhibit a mitigating effect. Additionally, our study explores the indirect impact of these factors using a panel threshold regression model. Results indicate a diminishing influence of economic growth on carbon emissions. Intriguingly, natural resource rents initially curtail, then amplify the connection between economic growth and carbon emissions. Conversely, rising energy intensity magnifies the relationship between economic expansion and carbon emissions. • Linear and nonlinear models were employed to explore the carbon emission influencing factors. • Explored the direct and indirect effects of geopolitical risk, natural resource rents, corruption and energy intensity on carbon emissions. • Natural resource rents, corruption governance in favour of lower carbon emissions, geopolitical risk and energy intensity in contrast. • Natural resource rents, corruption and geopolitical risks affect the relationship between economic growth and carbon emissions. [ABSTRACT FROM AUTHOR]