In this work, we introduce an enhanced square-root information filter for visual-inertial odometry. This filter utilizes stochastic cloning, implemented via Gaussian elimination, to facilitate time offset calibration and feature anchor changes. By using single-precision numbers within the filter, we significantly reduce computational load and memory requirements. In addition, we employ a fast Mahalanobis distance test and block Householder triangulation to accelerate the calculations. To mitigate feature drift from frame-to-frame optical flow, we create keyframes at regular intervals and refine long-tracked features between them. We use affine optical flow to compensate for patch deformations induced by possible large spatial transformations between keyframes. An analytical approach to computing the affine transformation is proposed. Experiments conducted on real-world data show that the proposed method achieves state-of-the-art performance at a much faster speed.

Knowing the relative rotation angle improves relative pose estimation accuracy. We consider the problem of computing relative motion from a non-minimal number of correspondences with a known relative rotation angle. While several solvers for minimum correspondences have been proposed, no non-minimal solver for this problem currently exists. In this work, we propose two non-minimal solvers for this problem. The first solver solves the problem using convex relaxation and semidefinite programming, yielding certifiable solutions. The second method approaches the problem through local eigenvalue optimization with random initialization. Increasing the number of initial guesses lowers the chances of missing the correct solution. We conduct experiments on synthetic and real data, confirming our methods' advantages over competing methods.

Finding relative pose for cameras is of vital importance in computer vision and robotics. We investigate the problem of relative motion estimation between successive frames from a minimal number of correspondences. Existing approximated methods use a first-order approximation to relative pose in order to simplify the problem and produce an estimate quickly. Our solution uses Cayley parameterization to represent rotation and simplifies the high-degree polynomials only at the very end of the formulation, resulting in more accurate models. Furthermore, our method can be more effective if the camera rotates mainly around one coordinate axis. By treating the main rotation component as the hidden variable in the solution, we can retain more high-degree terms for the main rotation part, considerably widening the effective approximation range. Our experiments show that our method is more accurate than existing approximated solver, and that it is still effective for relatively large motions. Besides, our method produces far fewer solutions than essential matrix parameterized solvers.

Implementing federated learning (FL) algorithms in wireless networks has garnered a wide range of attention. However, few works have considered the impact of user mobility on the learning performance. To fill this research gap, we develop a theoretical model to characterize the hierarchical federated learning (HFL) algorithm in wireless networks where the mobile users may roam across edge access points (APs), leading to incompletion of inconsistent FL training. We provide the convergence analysis of conventional HFL with user mobility. Our analysis proves that the learning performance of conventional HFL deteriorates drastically with highly-mobile users. And such a decline in the learning performance will be exacerbated with small number of participants and large data distribution divergences among users’ local data. To circumvent these issues, we propose a mobility-aware cluster federated learning (MACFL) algorithm by redesigning the access mechanism, local update rule, and model aggregation scheme. We also conduct experiments to evaluate the learning performance of conventional HFL, a cluster federated learning (CFL) with simple averaging, and our proposed MACFL. The results show that our MACFL can enhance the learning performance, especially for three different cases: ($i$) the case of users with non-independent and identically distributed (non-IID) data, ($ii$) the case of users with high mobility, and ($iii$) the case with a small number of users.

Recently, the implementation of federated learning (FL) in wireless networks becomes a hotspot due to its flexible collaborative learning methods and privacy-preserving benefits. However, most of the existing works overlook the impact of user mobility on the learning performance, which is critical. Specifically, the mobile users may roam among multiple edge access points (APs) during the local training procedures, leading to incompletion of inconsistent FL training. In this paper, we theoretically study the impact of user mobility on the FL in hierarchical wireless networks. In our system model, the network consists of one cloud server, several edge APs, and multiple mobile users that have their positions vary over time. During the local training process, users may stay in or move out of the coverage area of the originally attached edge AP. In such a practical context, we analyze the convergence rate of the FL algorithm and provide experiments to evaluate the learning performance under different network parameters. Our results provide insights in further improvements of FL in hierarchical wireless networks.