Cross-View Geo-Localization is essential for drone visual localization and navigation, which aims at establishing correlation between images collected by unmanned aerial vehicle (UAV) and satellite platforms in the same geographic area. Drastic changes in the drone’s viewpoints pose a significant challenge for methods based on image representation mining. Previous studies attempt to learn fine-grained image appearance features from various perspectives; however, they tend to underutilize the various state information of the UAV. This paper proposes a novel multimodal framework, CGSI (Context-Guided and UAV’s Status Informed), which leverages UAV state textual descriptions to mitigate scene bias caused by viewpoint differences. The following two issues are addressed to achieve more accurate and reliable multimodal geo-localization: 1) The domain gap across different datasets caused by the fixed UAV altitudes. We propose a Context-Guided Multimodal Tokenizer, which learns contextual vectors from multi-altitude visual features and utilizes them as adaptive text tokens. 2) Multimodal features are susceptible to state-feature ambiguity. We propose a Drone Group Graph Attention method to enhance the association between UAV visual feature with the same location ID but different states and exploit the intrinsic relationships to extract discriminative multimodal features. Extensive experiments on the University-1652 and SUES benchmark demonstrate that our CGSI significantly outperforms existing algorithms, achieving state-of-the-art performance. The substantial improvements observed in cross-region ablation experiments further showcase the superior domain generalization capability of our method.

The unmanned aerial vehicle (UAV) equipped with a multi-degree-of-freedom robotic arm can perform aerial grasping and contact-based aerial operations. This study proposes a centralized control strategy with a neural network-based disturbance estimator to achieve simultaneous precise control of both the manipulator and UAV. The RBF neural network is used to estimate the parametric uncertainties and external disturbances, which can enhance the composite system control performance without altering the original control architecture. To accelerate the system response speed, eliminate the steady-state errors, and simplify the control design, a fractional-order sliding mode control (FOSMC) framework is introduced to incorporate the fractional-order differential equations to precisely characterize the dynamic responses with high sensitivity and resolution, thereby improving the transient behavior. Further, a fast terminal sliding mode reaching rule is integrated to suppress the high-frequency chattering and enhance the control precision, ensuring finite-time convergence and robust steady-state characteristics. The system stability is validated through Lyapunov theory and physical flight experiments.

With the rapid advancement of autonomous driving technology, environmental perception, as its core component, faces critical challenges such as under low-illumination conditions. In this paper, a low-illumination environmental perception model based on YOLOv11 (You Only Look Once version 11) is proposed to tackle the performance degradation of visual sensors under low-illumination conditions. First, a progressive illumination enhancement module (PIE) is designed and incorporated into the backbone network of YOLOv11 model to capture the incremental local illumination. Then a histogram attention mechanism (HSSA) is introduced to address the illumination consistency issues, by capturing the illumination variations at different granularities. Finally, GSCONV is integrated into the model to implement lightweight improvements. The improved model can achieve a 7.1% increase in detection accuracy over the baseline model on the public dataset.

Visual place recognition (VPR) has played an essential role in simultaneous localization and mapping-based mobile robotics and autonomous driving in the past decade, which can identify previously visited places by matching the current observed view against a view database for global localization and loop closure. However, existing VPR methods always suffer fromfalse place recognitiondue to the following issues: insensitive spatial–temporal embedding extraction, lack of multiview descriptor for matching, and misconsideration of false views for model optimization. To address these issues, a novel robust framework calledVPR under variational views (VPR-VV)is proposed. VPR-VV is integrated with: a sequence encoder to extract robust spatial–temporal features from a view sequence, then a hierarchical view retrieval module is employed for multiview feature descriptor aggregation, and a novel enhanced ranking feedback average precision loss with the normalized discounted cumulative gain metric is designed for model optimization. As a result, VPR-VV can significantly enhance the accuracy and robustness of VPR for robot localization under variational views. Experiments with ablation studies are conducted on various challenging indoor and outdoor datasets, and our framework’s superiority is demonstrated: VPR-VV outperforms state-of-the-art (SOTA) methods by up to 9.4% in recall@1, and real-time inference is achieved without additional memory or computational overhead.

Accurate localization and pose estimation remain challenging for autonomous robots in low-texture environment. This article proposes a tightly coupled direct depth-inertial odometry and mapping (DDIO-Mapping) framework to simultaneously tackle three crucial issues in such environments: 1) ineffective feature point extraction; 2) inefficient searching of feature points; and 3) imbalanced feature extraction under uneven illumination conditions. In DDIO-Mapping, a novel robust strategy is designed that combines grayscale and depth features for optimization instead of only the RBG features in the existing methods. To improve searching efficiency, a new RGBD feature extraction is applied to directly extract both the depth and grayscale features from the RGBD images, which only requires searching the feature points in the 2-D space rather than the enormous 3-D space in K-dimensional (KD) tree. To deal with imbalanced feature extraction, a feature filtering and selection strategy is proposed to adaptively adjust the depth and grayscale weightage. Finally, with the effectively extracted features from RGBD images, a new nonlinear tightly coupled inverse depth residual function is customized to accurately estimate the optimal pose in low-texture environments. The framework is highly robust, accurate, and efficient. Experiments demonstrate that DDIO-Mapping reduces the root-mean-square error by approximately 30% compared to other state-of-the-art algorithms while retaining the same efficiency of approximately 20–35 ms.

Timely and effective search and rescue (SAR) is highly desired in the disaster rescues. Unmanned aerial vehicles (UAVs) can quickly conduct aerial searches to assist SAR with the equipped sensors. A visible-thermal human detection model based on an improved you only look once version 5 (YOLOv5) network is proposed to compensate the deficiencies of visible data with thermal images. The complementary information between the visible images and thermal images are considered with a partially shared two-stream backbone network, so as to better preserve the information of each branch while reducing the domain distinction and extracting the modality-invariant features. Features of the two modalities are fused via a fusion module with a multidimensional attention mechanism. By taking the pixels outside the region of interest as negative samples, the extra loss function can suppress the uncorrelated feature extraction of the backbone to enhance the effective feature representation. The proposed visible-thermal human detection model has been deployed on the UAV with satisfied human detection performance. Comparative experiments on the multispectral pedestrian dataset KAIST have also been performed to demonstrate that the proposed model outperforms other visible-thermal object detection models with log-average miss rate.

Multiview multi-instance multilabel learning (M3L) is a popular research topic during the past few years in modeling complex real-world objects such as medical images and subtitled video. However, existing M3L methods suffer from relatively low accuracy and training efficiency for large datasets due to several issues: 1) the viewwise intercorrelation (i.e., the correlations of instances and/or bags between different views) are neglected; 2) the diverse correlations (e.g., viewwise intercorrelation, interinstance correlation, and interlabel correlation) are not jointly considered; and 3) high computation burden for training process over bags, instances, and labels across different views. To resolve these issues, a novel framework called fast broad M3L (FBM3L) is proposed with three innovations: 1) utilization of viewwise intercorrelation for better modeling of M3L tasks while existing M3L methods have not considered; 2) based on graph convolutional network (GCN) and broad learning system (BLS), a viewwise subnetwork is newly designed to achieve joint learning among the diverse correlations; and 3) under BLS platform, FBM3L can learn multiple subnetworks jointly across all views with significantly less training time. Experiments show that FBM3L is highly competitive (or even better than) in all evaluation metrics [up to 64% in average precision (AP)] and much faster than most M3L (or MIML) methods (up to 1030 times), especially on large multiview datasets (≥260 K objects).

In the process of grasping objects with an unmanned aerial manipulator (UAM), the weight of the payload is not always known in advance. Using model-based high-performance controllers to capture unknown objects introduces new interference to the system, which has a negative impact on the closed-loop control performance. This paper proposes a trajectory tracking controller for UAM during the object grasping with unknown mass. An adaptive module using back-stepping is developed for online estimation of the object mass, and an improved super-twisting extended state observer (ISTESO) is designed to cope with external disturbances and counteract uncertainties. Lya-punov stability theory is used for stability analysis to ensure that the trajectory tracking converges with relative estimation error approaching zero. Simulation experiments have been performed to verify the effectiveness of the proposed controller.

Label distribution learning (LDL), leveraging the label significance (LS), is more appropriate for solving label ambiguity problems than multilabel learning (MLL). However, directly obtaining the LS of LDL is extremely expensive and challenging. Thus, label enhancement (LE) algorithms are effectively proposed to acquire inherent LS from MLL for training the LDL models. Nevertheless, most existing LE models will suffer from low accuracy and low efficiency with following issues: ignoring mapping relationship between feature and label space, resulting in inaccurate enhanced data; designing independently apart from LDL models, resulting in an inability of unified LE-LDL learning; and requiring to optimize numerous parameters iteratively, resulting in worse training efficiency. Consequently, a novel unified LE-LDL learning framework, namely stacked graph-regularized polynomial-based fuzzy broad learning system (SGP-FBLS), is proposed by following three innovations: polynomial-based fuzzy system is introduced to enhance feature mapping ability while improving the learning performance effectively; graph regularized-based optimization objective function (GP-FBLS) is presented by considering interinstance correlation and label correlation to mine potential LS, thereby improving the accuracy of subsequent LDL tasks; and a weight stacked strategy is innovatively proposed to directly transmit LS and weighted parameters from GP-FBLS to SGP-FBLS without retraining, achieving the most satisfactory performance while significantly improving the training efficiency. Finally, comparative studies on 19 practical datasets demonstrate the effectiveness and superiority of proposed methods.

Multi-label learning for large-scale data is a grand challenge because of a large number of labels with a complex data structure. Hence, the existing large-scale multi-label methods either have unsatisfactory classification performance or are extremely time-consuming for training utilizing a massive amount of data. A broad learning system (BLS), a flat network with the advantages of succinct structures, is appropriate for addressing large-scale tasks. However, existing BLS models are not directly applicable for large-scale multi-label learning due to the large and complex label space. In this work, a novel multi-label classifier based on BLS (called BLS-MLL) is proposed with two new mechanisms: kernel-based feature reduction module and correlation-based label thresholding. The kernel-based feature reduction module contains three layers, namely, the feature mapping layer, enhancement nodes layer, and feature reduction layer. The feature mapping layer employs elastic network regularization to solve the randomness of features in order to improve performance. In the enhancement nodes layer, the kernel method is applied for high-dimensional nonlinear conversion to achieve high efficiency. The newly constructed feature reduction layer is used to further significantly improve both the training efficiency and accuracy when facing high-dimensionality with abundant or noisy information embedded in large-scale data. The correlation-based label thresholding enables BLS-MLL to generate a label-thresholding function for effective conversion of the final decision values to logical outputs, thus, improving the classification performance. Finally, experimental comparisons among six state-of-the-art multi-label classifiers on ten datasets demonstrate the effectiveness of the proposed BLS-MLL. The results of the classification performance show that BLS-MLL outperforms the compared algorithms in 86% of cases with better training efficiency in 90% of cases.

In this work, we focus on real-time 3D reconstruction or localization and mapping for outdoor scene using an aerial vehicle called quadcopter. Quadcopter provides the advantages of high flexibility and wide view field in spatial movement. However, existing feature-based and direct methods (using dense or semi-dense approach) are not suitable for outdoor environment, in which multiple challenging scenarios arise such as lighting variance, jittering views, high-speed and non-smooth flight trajectory. The main reason is that the existing methods rely on the assumption of brightness constancy across multiple images and only raw pixel intensities are employed for direct image alignment. In order to tackle these scenarios, a novel method called Feature-based Direct Tracking and Mapping (FDTAM) is proposed, which i) incorporates an efficient binary feature descriptor into direct image alignment module to tackle the challenging scenarios, such as drifting issue under lighting variance problem; ii) applies semi-dense approach to obtain high reconstruction quality; iii) provides a framework with low computational complexity for real-time reconstruction. Compared to other state-of-the-art feature-based and direct methods, our proposed method is shown to tackle the challenging scenarios and improve the accuracy and robustness even in CPU (rather than GPU) platform.

Knowledge distillation is one of the model compression techniques, where the key challenge is how to extract rich and general knowledge from teacher models, thus reducing the performance gap between the student models and the teacher models. The feature representation of the strong convergence of the teacher models has a heavy constraint on the student models. which brings certain influence on the quality of the student models. This paper designs an online knowledge distillation based on multi-stage multi-generative adversarial network (MS-MGAN), which enables the student model and the teacher model to learn together in multiple stages. First, the teacher model training is divided into several stages, and the training results of each stage are used to guide the student model training. Second, the student model is based on the convolutional block for layer-wise greedy training. Then, the generative adversarial network is integrated into the multi-stage training, with each convolutional block of the student model as the generator and the corresponding block of the teacher model as the discriminator, so that the student model can better follow and even surpass the performance of the teacher model. As demonstrated on several benchmarks, the proposed method can achieves surprising satisfied results, even outperforming the teacher model on some tasks.

When a small portion of the decision makers hold the correct information and the majority hold the opposite, the correct ranking of the alternatives for the group decision-making cannot be obtained with the current methods. A novel method is thus developed to tackle this challenge in this article. The priori probabilities of each alternative can be calculated via the opinions of the group decision makers, which are presented as the pairwise comparisons of the alternatives in the form of the linguistic preference relation. Based on the aggregated probabilities of the alternatives in the group of the decision makers, the normalized-prediction selection rate (NPSR) is defined and calculated accordingly. The alternative with maximal NPSR is selected as the correct answer, whereas the accuracy of the correct alternative selection (CAS) is guaranteed by two propositions. The iterative algorithm is first devised to determine the ranking of the alternatives depending on the CAS. For the proposed method, the decision makers require no modification of the opinions as can avoid the consensus problem, and the CAS can be obtained under the circumstances that the correct information is held by the minority of the group. Finally, the experiment has been conducted to demonstrate the efficacy of the proposed method to obtain the CAS, and the main limitations of proposed method are carefully addressed as well.

In this paper, the walking gait of the biped robot is studied. An optimized method for the trajectory of the swinging legs based on Bezier curve is proposed. First, the impact of the trajectory of the swinging legs on the stability of the biped robot is analyzed. The Bezier based curve is then used for trajectory planning of the end-leg. Through the Newton tangent method, the interpolation points of the swinging leg with uniform movement can be solved with the aid of the Bezier curve. So the biped robot can maintain the uniform motion and reduce the impact of inertial force during the process from acceleration to deceleration. Further, the biped robot can move in fast speed to avoid falling and improve the walking stability. Experiments are performed in Matlab simulation and Alpha Ebot. The desired trajectory of each joint of the biped robot is planned and applied to the real robot platform to verify the effectiveness of the optimized gait planning method.

In this paper, a novel active disturbance rejection controller (ADRC) of ship steering control is designed under the internal parameter uncertainties and external disturbances. The ship heading control model is rather a complex model with large time delay, uncertainties and serious external disturbances, which may result in poor control performance. Therefore, ADRC is applied to control the steering of the ship. However, manually tuning the parameters of the ADRC is timeconsuming and labor-intensive. An optimization method based on adaptive particle swarm optimization (APSO) to adjust the ship’s steering ADRC parameters is proposed. In the developed APSO, considering the control performance, the fitness function is modified in the light of the integrated time absolute error (ITAE) criteria. Then the inertia weight of the previous velocity in the current velocity update is varied based on the fitness value of each individual at the previous moment. Finally, the effectiveness of the proposed adaptive optimization method is proved by simulation experiments.

Learning from Demonstration (LfD) has been identified as an effective method for making robots adapt to a similar kind of tasks. In this work, a framework of learning from demonstration has been proposed for modelling robot motions. We present an approach based on dimension ascending to learn a dynamical system, so that the reproduced motions can closely follow the demonstrations. In addition, the reproductions can ultimately reach and stop at the target, which reflects the robustness of the method. Therefore, the system accuracy and stability can be better guaranteed simultaneously. The effectiveness of the proposed approach is verified by performing handwriting experiments on the LASA data set.

Economic unit commitment is a mix-integer large scale optimization problem calling for powerful and efficient tools. On the other hand, environmental impact related to the power generation is attracting increasing attentions due to the global warming trend and urgent calls for sustainable energy development. In this paper, the dual objectives of economic and emission unit commitment is converted into a single objective problem. For solving this, a novel binary Jaya optimization is proposed and integrated with lambda iteration method. The proposed binary Jaya method is inspired by the Jaya evolution and generates binary bits from a v-shape transfer function. Numerical study demonstrates the significant improvement of the binary Jaya in regarding the convergence speed for solving unit commitment problem. The solution distributions of the both objectives also show the effective of the proposed methods.

In this paper, a novel attitude estimation for unmanned aerial vehicle (UAV) is proposed based on long and short term memory neural network (LSTM NN). The UAV is a strong coupling and multi-variable nonlinear complex system, in which the attitude estimation is nonlinear and the attitude data of the UAV is a time series sequence. LSTM NN is therefore selected due to its satisfied performance in time-based data prediction. The data samples to train the LSTM NN are collected during the test flight of a quadrotor. To improve the accuracy of the model, different configurations of the LSTM NNs are used for comparison. Experimental results demonstrate that the method for the UAV attitude estimation has higher accuracy and the potential of applying deep learning technique to the online UAV attitude estimation.

The drones formation control in 5G wireless sensor network is discussed. The base station (BS) is used to receive backhaul position signals from the lead drone in formation and launches the beam to the lead one as the fronthaul flying signal enhancement. It is a promising approach to raise the formation strength of drones during flight control. The BS can transform the direction of the antennas and transmit energy to the lead drone that could widely enlarge the number of the receivers and increase the transmission speed of the data links. The millimeter‐Wave (mmWave) communication system offers new opportunities to meet this requirement owing to the tremendous amount of available spectrums. However, the massive non‐line‐of‐sight (NLoS) transmission and the site constraints in urban environment are severely challenging the conventional deploying terrestrial low power nodes (LPNs). Simulation experiments have been performed to verify the availability and effectiveness of mmWave in 5G wireless sensor network.

In this paper, a method is proposed to search for spatio-temporal path for action localization in unconstrained videos. We mainly focus on two requirements, i.e., accurate human extraction and speeding generation of action proposal. The approach first generates human proposals at the frame level, then scores them based on two complementary parts, i.e., posteriori probability evaluated via a fine-tuned Faster-RCNN and template-matching similarity based on the spatiotemporal continuity. Finally, the generation of action proposal is formulated as a Max-Path discovery problem, coupled with dynamic programming to find an optimal path with maximum score. Experiments on UCF-Sports are performed to verify that the proposed method can achieve fast high-quality action proposal and link the missed-detection proposals in successive frames together to form a complete action.

In this paper, nuclear radiation distribution (NRD) prediction is discussed and modelled. Firstly, the nuclear radiation distribution model in static state is studied, and the leakage source is regarded as a continuous leak point and proliferating around in evenly speed. The density of the NRD is calculated under the conditions of wind, ground reflection, particle sedimentation and nuclear decay. A detection system is developed around the nuclear plant, including nuclear radiation detector, node of temperature and humidity, wind direction and speed node. The prediction of the NRD is used to detect the densities, where the estimation values are used for comparison in close distance with the measurements obtained directly from the detector and used for estimation in long distance (> 300m). Simulation experiments are performed to prove the accuracy of the proposed distribution prediction.

This paper proposed a novel image processing method combining Principal Component Analysis (PCA) and Genetic Algorithm (GA) to reduce the interference of facial expression, lighting or wear but extracting gender feature from frontal face. The collected facial images are first cropped and aligned automatically, then the gray-level information can be converted to feature vectors via PCA. After eigen-features are extracted with high classification performance by the aid of GA, the neural network classifier can be trained accordingly. Compared to the classification methods based on global gray-level information, the obtained classifier has better identification rate but half less used feature dimension, so the calculation load can substantially be reduced during training and identification procedures, which benefits to the development of a real-time identification system. Furthermore, FERET dataset and FEI dataset are used to validate the generality of the proposed method, where 92% and 94% accuracy rates of the gender recognition can be achieved respectively.

In this paper, a method to predict the power charging demand and discharging output of the electric vehicles (EVs) is proposed. Besides EVs are the energy end-users powered by batteries, they can also be regarded as distributed storage units to participate the energy scheduling in the power grid. The appearance of V2G technologies provides an opportunity for energy dual connection and communication, which can lower the cost of EVs, improve power flexibility and reduce economical operation cost of grids. Monte carlo method is applied to predict the power demand of EVs in real-time so as to regulate the EVs charging or charging patterns. Matlab is used as a simulation tool to perform the experiments to estimate the power demand and output related to EVs.

It is known that, in the current hybrid video coding structure, spatial and temporal prediction techniques are extensively used which introduce strong dependency among coding units. Such dependency poses a great challenge to perform a global rate-distortion optimization (RDO) when encoding a video sequence. RDO is usually performed in a way that coding efficiency of each coding unit is optimized independently without considering dependeny among coding units, leading to a suboptimal coding result for the whole sequence. In this paper, we investigate the inter-frame dependent RDO, where the impact of coding performance of the current coding unit on that of the following frames is considered. Accordingly, an inter-frame dependent rate-distortion optimization scheme is proposed and implemented on the newest video coding standard High Efficiency Video Coding (HEVC) platform. Experimental results show that the proposed scheme can achieve about 3.19% BD-rate saving in average over the state-of-the-art HEVC codec (HM15.0) in the low-delay B coding structure, with no extra encoding time. It obtains a significantly higher coding gain than the multiple QP (±3) optimization technique which would greatly increase the encoding time by a factor of about 6. Coupled with the multiple QP optimization, the proposed scheme can further achieve a higher BD-rate saving of 5.57% and 4.07% in average than the HEVC codec and the multiple QP optimization enabled HEVC codec, respectively.

This paper addresses the robust H∞filtering problem for a class of Markov jump Lur'e systems with time-varying transition probabilities in discrete-time domain. The time-varying character of transition probabilities is considered to be finite piecewise homogeneous. A full-order filter is designed such that the resulting closed-loop systems are stochastically stable and have a guaranteed H∞performance index in terms of linear matrix inequalities. The effectiveness and potential of the developed results are verified through an example about a class of economic systems.

In this paper, an indoor navigation algorithm is proposed for the purpose of robot autonomous path planning. Due to the complex situation in indoor environments, it can cause a serious trouble for robot to identify the route during patrolling, especially for corner and door detection, which is the key step for intelligent navigation. To solve this problem, a kinect sensor is used for the door detection and corner location via depth images. The continuously varied ratios and depth difference in the images have been analyzed for the corner and door identification. Furthermore, the precise position of the doors and corners can be localized via the 3-dimensional characteristics of the depth images. Experiments in different scenarios have been performed to verify the efficacy of the algorithm for robot indoor autonomous navigation.