Search

Domain generalization (DG) aims to learn knowledge from multiple related domains to achieve a robust generalization performance in unseen target domains, which is an effective approach to mitigate domain shift in remote sensing image classification. Although the sharpness-aware minimization (SAM) method enhances DG capability and improves remote sensing image classification performance by promoting the convergence of the loss minimum to a flatter loss surface, the perturbation loss (maximum loss within the neighborhood of a local minimum) of SAM fails to accurately measure the true sharpness of the loss landscape. Furthermore, its variants often overlook gradient conflicts, thereby limiting further improvement in DG performance. In this paper, we introduce implicit sharpness-aware minimization (ISAM), a novel method that addresses the deficiencies of SAM and mitigates gradient conflicts. Specifically, we demonstrate that the discrepancy in training loss during gradient ascent or descent serves as an equivalent measure of the dominant eigenvalue of the Hessian matrix. This discrepancy provides a reliable measure for sharpness. ISAM effectively reduces sharpness and mitigates potential conflicts between gradients by implicitly minimizing the discrepancy between training losses while ensuring a sufficiently low minimum through minimizing perturbation loss. Extensive experiments and analyses demonstrate that ISAM significantly enhances the model's generalization ability on remote sensing and DG datasets, outperforming existing state-of-the-art methods. [ABSTRACT FROM AUTHOR]

The ternary chalcogenide AgBiS2 has attracted widespread attention in the field of photovoltaic and photoelectric devices due to its excellent properties. In this study, AgBiS2 powders with an average diameter of 200 nm were prepared via a simple and convenient reflux method from silver acetate, bismuth nitrate pentahydrate, and n-dodecyl mercaptan. The adjustment of the ratios of Ag:Bi:S raw materials and of the reaction temperatures were carried out to investigate the significance of the synthesis conditions toward the composition of the as-synthesized AgBiS2. The results of XRD indicated that the powders synthesized at a ratio of 1.05:1:2.1 and a synthesis temperature of 225 °C have the lowest bismuth content and the highest purity. The synthesized AgBiS2 crystallizes in a rock salt type structure with the cubic F m 3 ¯ m space group. Scanning and transmission electron microscopy, thermogravimetric analysis, ultraviolet–visible–near-infrared spectra, and photocatalytic degradation performance were employed to characterize the as-synthesized samples. The results demonstrated that AgBiS2 powders display thermal stability; strong absorption in the ultraviolet, visible, and partial infrared regions; and an optical bandgap of 0.98 eV. The obtained AgBiS2 powders also have a good degradation effect on the methylene blue solution with a degradation efficiency of 58.61% and a rate constant of 0.0034 min−1, indicating that it is an efficient strategy for sewage degradation to reduce water pollution. [ABSTRACT FROM AUTHOR]