This work investigates Unsupervised Domain Adaptation (UDA)-based semantic segmentation of Very High Resolution (VHR) remote sensing images from different domains. Most existing UDA methods resort to Generative Adversarial Networks (GAN) to cope with the domain shift problem caused by the discrepancies across different domains. However, these GAN-based UDA methods directly align two domains in the appearance, latent, or output space based on convolutional neural networks, making them ineffective in exploiting long-range dependencies across the high-level feature maps derived from different domains. Unfortunately, such high-level features play an essential role in characterizing remote sensing images with complex content. To circumvent this obstacle, Mutually Boosted Attention Transformer (MBATrans) is proposed to capture cross-domain dependencies of semantic feature representations in this work. Compared with conventional UDA methods, MBATrans can significantly reduce domain discrepancies by capturing transferable features using global attention. More specifically, MBATrans utilizes a novel Mutually Boosted Attention (MBA) module to align cross-domain feature maps while enhancing domain-general features. Furthermore, a novel GAN-based network with improved discriminative capability is devised by integrating an additional discriminator to learn domain-specific features. Extensive experiments on two large-scale VHR remote sensing datasets, namely ISPRS Potsdam and Vaihingen, confirm the superior performance of the proposed MBATrans-Augmented-GAN (MBATA-GAN) architecture. The source code in this work is available at https://github.com/sstary/SSRS.

While many distributed optimization algorithms have been proposed for solving smooth or convex problems over the networks, few of them can handle non-convex and non-smooth problems. Based on a proximal primal-dual approach, this paper presents a new (stochastic) distributed algorithm with Nesterov momentum for accelerated optimization of non-convex and non-smooth problems. Theoretically, we show that the proposed algorithm can achieve an ϵ-stationary solution under a constant step size with O (1/ϵ ² ) computation complexity and O (1/ϵ) communication complexity when the epigraph of the non-smooth term is a polyhedral set. When compared to the existing gradient tracking based methods, the proposed algorithm has the same order of computation complexity but lower order of communication complexity. To the best of our knowledge, the presented result is the first stochastic algorithm with the O (1/ϵ) communication complexity for non-convex and non-smooth problems. Numerical experiments for a distributed non-convex regression problem and a deep neural network based classification problem are presented to illustrate the effectiveness of the proposed algorithms.

This paper considers the set membership fusion estimation problem for the general multisensor dynamic systems with unknown but bounded noises. In order to achieve the ellipsoidal fusion formulae, we divide the set membership filter problem into two steps: the prediction step and the fusion update step. The proposed method has the following nice properties. Firstly, each step can be converted into a semidefinite programming problem which can be efficiently computed. Secondly, partanalytical formulae of the shape matrix and the center of the bounding ellipsoid can be derived by using a decoupled technique, which can significantly reduce the computational complexity. Moreover, the relationship between the fusion center and local sensors can be clearly revealed based on the ellipsoidal fusion formula. Finally, it is interesting that the ellipsoidal fusion formula is similar in form to the classic mean squared error filter fusion, although they are of the different optimization frameworks. However, the fusion weights of the information matrix provided by the different sensors are different and the optimal fusion weights can be calculated by solving a convex optimization problem. A typical numerical example in target tracking demonstrates the effectiveness of the proposed centralized and distributed fusion formula

In this article, we propose a dual set membership filter for nonlinear dynamic systems with additive unknown but bounded noises, and it has three distinct advantages. First, the nonlinear system is translated into the linear system by leveraging a semi-infinite programming, rather than linearizing the nonlinear function. The semi-infinite programming is to find an ellipsoid bounding the nonlinear transformation of an ellipsoid, which aims to compute a tight ellipsoid to cover the state. Second, the duality result of the semi-infinite programming is derived by rigorous analysis; then, a first-order Frank–Wolfe method is developed to efficiently solve it with a lower computation complexity. Third, the proposed filter enjoys stability for some special nonlinear dynamic systems and succeeds the advantages of the classic linear set membership filter. Finally, two illustrative examples in the simulations reveal the effectiveness of the dual set membership filter.

In this paper, we propose a MIMO transmit beamforming optimization model for joint radar sensing and multi-user communications, where the design of the beamformers is formulated as an optimization problem whose objective is a weighted combination of the sum rate and the Cramér-Rao bound, subject to the transmit power budget. Obtaining the global solution for the formulated nonconvex problem is a challenging task, since the sum-rate maximization problem itself (even without considering the sensing metric) is known to be NP-hard. The main contributions of this paper are threefold. Firstly, we derive an optimal closed-form solution to the formulated problem in the single-user case and the multi-user case where the channel vectors of different users are orthogonal. Secondly, for the general multi-user case, we propose a novel branch and bound (B&B) algorithm based on the McCormick envelope relaxation. The proposed algorithm is guaranteed to find the globally optimal solution to the formulated problem. Thirdly, we design a graph neural network (GNN) based pruning policy to determine irrelevant nodes that can be directly pruned in the proposed B&B algorithm, thereby significantly reducing the number of unnecessary enumerations therein and improving its computational efficiency. Simulation results show the efficiency of the proposed vanilla and GNN-based accelerated B&B algorithms.

In this paper, we propose a tighter set-membership filter for some nonlinear dynamic systems by using analytic method and boundary sampling technique. The nonlinear dynamic systems can be linearized about the current estimate, then the remainder term is bounded in real time by an optimization ellipsoid, other than a priori remainder bound. For a twodimensional radar system and a quadratic system, some regular properties can be derived for the remainder term, which helps us obtain a tighter bounding ellipsoid to cover the remainder. Moreover, the prediction step and the measurement update step are derived based on the recent optimization method and the on-line bounding ellipsoid of the remainder, so that a tighter set-membership filter can be achieved. The numerical examples demonstrate the effectiveness of the proposed filter.