To enable unmanned aerial vehicle (UAV) operators to efficiently and intuitively convey their commands to a swarm of UAVs, we propose the use of natural and human-centric input modalities, such as voices and gestures. This paper addresses the fusion of input modalities such as voice and gesture data, which are captured through a microphone and a Leap Motion controller, respectively, to control UAV swarms. The obtained experimental results are presented, and the achieved performance (accuracy) is analyzed. Finally, combined human factor ergonomics test with a questionnaire to verify the method’s validity.

This work proposes a geometric Covariance Intersection (CI) based algorithm for efficiently solving the decentralized Cooperative Localization (CL) problem with multiple TOA measurements. Unlike existing algorithms that only process one measurement each time, our proposed algorithm processes multiple measurements in a batch style, which improves the precision performance by considering the correlation among these measurements. An efficient way to get the optimal solution is given in our algorithm, which reduces the computational complexity to O(NlogN). Furthermore, the sparsity of the optimal solution is proved, which enables a heuristic state buffering approach to reduce the usage of ranging link resources and improve measurement efficiency. Simulation results show that our algorithm outperforms the conventional methods.

Image-based ground-to-aerial geo-localization aims to determine the geo-location of a ground query image by matching it with a large geo-tagged aerial image database. Due to the drastic difference between ground and aerial views, achieving high-accuracy geo-localization remains a huge challenge, especially in practical scenarios where ground query images have unknown orientations and even limited field-of-views (FoVs). Incomplete information significantly hampers the process of learning discriminative features for image matching. In this paper, we propose a novel AutoMatic Progressive LEarning method (AMPLE) for the orientation unknown geo-localization task. Specifically, we design a ConvNeXt-based network to effectively extract orientation-aware features from the two views. We then present two progressive training strategies without manually predefined training stages to promote the learning process. The first adaptively mines harder negative samples that contribute more to the loss, by automatically discarding redundant samples as the current best accuracy increases. The second leverages the proposed alignment-correlation hybrid loss to guide model optimization in a progressive manner, gradually reducing the reliance on auxiliary orientation information. Extensive experiments on two benchmark datasets demonstrate that AMPLE outperforms state-of-the-art methods in orientation unknown, FoV limited, and cross-area tasks. Lastly, we propose the concept of unifying unknown orientation tasks at different FoVs, and show the cross-FoV generalization capability of our method.

In this letter, we study the localization problem of a mobile robot with range measurement from a single beacon. Previous filtering-based studies usually required accurate statistics of noises to be theoretically sound and reliable, which is difficult to obtain in practical systems. To solve this problem, we propose an accurate and efficient localization method based on a set-membership filtering framework with constrained zonotopes. This method includes three novel steps. First, we design a convex optimization relaxation method to handle the non-convexity caused by the single-beacon range measurement. Then, a halfspace-intersection refinement is proposed which improves the estimation accuracy. Finally, we provide a sliding-window recursive method that simultaneously guarantees the accuracy and the efficiency. Simulations and field experiments corroborate the effectiveness of our proposed method.

Equipped with inter-satellite links (ISLs), BeiDou-3 satellites are able to operate autonomously in case of ground facility malfunction. To provide reliable positioning, navigation and timing service during autonomous operation, satellites need to maintain a stable time reference. In this letter, we propose a minimal covariance distributed Kalman filter (MCDKF) to realize distributed timekeeping. The proposed algorithm overcomes the problem of network dynamics. Meanwhile, it improves the stability of the on-board time reference by minimizing the variance of the estimate errors at each update step, under the constraint of local observations. Simulation results show that the proposed algorithm reduces the estimation variance and improves the frequency stability compared to other distributed algorithms.