行人重识别虽已取得了显著进展,但在实际应用场景中,不同障碍物引起的遮挡问题仍然是一个亟待解决的挑战。为了从被遮挡行人中提取更有效的特征,提出了一种基于可学习掩模和位置编码(Learnable mask and position encoding, LMPE)的遮...行人重识别虽已取得了显著进展,但在实际应用场景中,不同障碍物引起的遮挡问题仍然是一个亟待解决的挑战。为了从被遮挡行人中提取更有效的特征,提出了一种基于可学习掩模和位置编码(Learnable mask and position encoding, LMPE)的遮挡行人重识别方法。首先,引入了一种可学习的双路注意力掩模生成器(Learnable dual attention mask generator, LDAMG),生成的掩模能够适应不同遮挡模式,显著提升了对被遮挡行人的识别准确性。该模块可以使网络更灵活,能更好地适应多样性的遮挡情况,有效克服了遮挡带来的困扰。同时,该网络通过掩模学习上下文信息,进一步增强了对行人所处场景的理解力。此外,为了解决Transformer位置信息损耗问题,引入了遮挡感知位置编码融合(Occlusion aware position encoding fusion, OAPEF)模块。该模块进行不同层次位置编码融合,使网络获得更强的表达能力。通过全方位整合图像位置编码,可以更准确地理解行人间的空间关系,提高模型对遮挡情况的适应能力。最后,仿真实验表明,本文提出的LMPE在Occluded-Duke和Occluded-ReID遮挡数据集以及Market-1501和DukeMTMC-ReID无遮挡数据集上都取得了较好的效果,验证了本文方法的有效性和优越性。展开更多
In this paper,we study the optimal investment problem of an insurer whose surplus process follows the diffusion approximation of the classical Cramer-Lundberg model.Investment in the foreign markets is allowed,and the...In this paper,we study the optimal investment problem of an insurer whose surplus process follows the diffusion approximation of the classical Cramer-Lundberg model.Investment in the foreign markets is allowed,and therefore,the foreign exchange rate model is incorporated.Under the allowing of selling and borrowing,the problem of maximizing the expected exponential utility of terminal wealth is studied.By solving the corresponding Hamilton-Jacobi-Bellman equations,the optimal investment strategies and value functions are obtained.Finally,numerical analysis is presented.展开更多
基金supported by the National Natural Science Foundation of China(Grant No.12301603).
文摘In this paper,we study the optimal investment problem of an insurer whose surplus process follows the diffusion approximation of the classical Cramer-Lundberg model.Investment in the foreign markets is allowed,and therefore,the foreign exchange rate model is incorporated.Under the allowing of selling and borrowing,the problem of maximizing the expected exponential utility of terminal wealth is studied.By solving the corresponding Hamilton-Jacobi-Bellman equations,the optimal investment strategies and value functions are obtained.Finally,numerical analysis is presented.
