Tashkent State University of Economics

In a bid to meet the challenging situations of facial expression recognition (FER) in real-time and with high accuracy, the researchers have come up with a radical solution: The Lightweight Adaptive Multi-Scale Fusion Transformer (AMFT). This new FER framework is a brilliant one that links multi-scale feature extraction with an adaptive fusion method, along with a state-of-the-art transformer architecture, to take the accuracy to a great level while maintaining its efficiency. Apart from the rest, AMFT has been conceptualized to use less computation, thereby making it a very promising real-time application that can be used in different areas such as security, healthcare, and interactive computing. The core of the innovation in the model is its feature of being able to dynamically select the degree of facial expression intricacy; thus, the energy to be used for processing different scenarios will be regulated without any compromise in performance. Performance of the system was measured on standard datasets, and it was found that AMFT outperforms the existing models by a wide margin; hence, it gives not only faster processing speeds but also lowers computational demands. The characteristic of the architecture that led to this feat is the combination of a fusion method that is both adaptive and based on multi-scale processing and the use of a transformer for further enhancements, which represents a breakthrough in FER technology, allowing easier transfer from laboratory experiments to real-world environments.

Abstract Purpose: This study aims to systematically analyze methods for determining the level of trust in the evaluation of the higher education system and to substantiate their scientific foundations by conceptualizing trust as a measurable and analytically significant dimension of evaluation Research Methodology:. A mixed-methods approach was employed by integrating quantitative statistical indicators and qualitative sociological survey data from key stakeholders, including students, academic staff, parents, and employers. Data were analyzed using descriptive statistics and expert-based weighting to construct an integrated trust index Results: The results show that trust in higher education evaluation is influenced by both system performance and stakeholders’ perceptions of fairness, transparency, clarity, and relevance. Graduate employment rates and academic staff capacity were the strongest quantitative determinants, while transparency and clarity of evaluation procedures were the most influential qualitative factors. Qualitative indicators slightly outweighed quantitative ones in shaping overall trust Conclusions: The study concludes that an integrated evaluation framework combining quantitative and qualitative indicators provides a more comprehensive assessment of trust and enhances the social legitimacy of higher education evaluation systems Limitations: The study is limited by its cross-sectional design, limited sample size, reliance on expert weighting, and the use of self-reported data Contribution: This research contributes by proposing an integrated trust index that can support policymakers and institutions in improving transparency, stakeholder confidence, and the effectiveness of higher education evaluation

The review is a well-structured, detailed study of Maxwell fluid flow across a wide range of geometries, including, but not limited to, stretching sheets, vertical plates, inclined plates, cylinders, and channels. One hundred and ninety-six articles were identified; after systematic screening, 133 pertinent studies were selected. It is reviewed under steady and unsteady flow conditions and investigates the impact of major physical effects, i.e., magnetohydrodynamics (MHD), thermal radiation, chemical reactions, porous media, and heat generation. However, unlike the prevailing literature, which focuses on individual configurations or parameters, this research offers a comparative and critical assessment by combining various geometries, physical effects, and solution methods. The analysis shows that most studies focus on simplified geometries and single-parameter effects. In contrast, the interaction of multiple physical processes and realistic configurations is not studied to the same extent. As found in the review, there is an evident gap in research on multi-physics interaction studies across various geometries. This work, by summarizing the available knowledge and highlighting the underexplored fields, provides a detailed outline of future research in Maxwell fluid dynamics with immediate implications for engineering and industry.

This study examines the behavior of compact astrophysical objects within a matter-geometry coupled f (R) gravity model. The modified field equations are expressed for a static interior spacetime with an anisotropic matter distribution. Applying two well-defined radial components of the metric ansatz and anisotropic pressures allows for analytical solutions to these equations. In both theoretical models, integrating the differential equations troduces constants, which are fixed using boundary conditions. Furthermore, the condition of null radial pressure at the boundary is used to determine these constants. Additionally, we visually assess certain important features that ensure the physical acceptability of the proposed model and support our analysis with observational from LMC X-4. Our theoretical research shows that both models meet the physical viability and stability requirements. Further, our investigation also contributes to the knowledge of how the modified gravity model influences the interior structure of compact stars, paving the way for future studies.

China's light industry is very important to global value chains, but we still don't know enough about how it contributes to embodied carbon emissions at the sector level. This research utilizes a global multi-regional input-output framework derived from the World Input-Output Database to measure both production-based and consumption-based embodied carbon emissions linked to China's light industrial exports from 2002 to 2016. The analysis uses the Leontief inverse matrix and sectoral carbon intensity coefficients to follow direct and indirect carbon transfers between 42 economies and 56 sectors. The results show that China is always a net exporter of embodied carbon emissions, especially in carbon-heavy industries like textiles, leather, and paper products. Production-based accounting assigns considerable emissions to domestic industrial activities, whereas consumption-based analysis reveals that a big portion of these emissions is influenced by international demand. This difference shows that there are structural imbalances in global carbon responsibilities and provides real-world proof of how carbon leakage works in international trade. To bolster robustness, the work employs cross-database validation with EXIOBASE and EORA, thereby affirming that WIOD-based estimations maintain consistency within acceptable analytical margins. The study connects embodied carbon flows to carbon leakage, comparative advantage in pollution-heavy industries, and environmental justice issues. This adds to the continuing discussions about how to make trade more sustainable and fair climate governance. The findings provide pertinent insights for sector-specific decarbonization, preparedness for carbon border adjustments, and synchronized global mitigation efforts amid the developing climate policy frameworks.