Unmanned Aerial Vehicles (UAVs) play a significant role in wireless communication because of their high maneuverability and the advantage of forming Line-of-Sight (LoS) links with ground users. In this research, we investigate the trajectory design and resource scheduling problem of UAV relay communication optimization scenarios, considering heterogeneous tasks and QoS (Quality of Service) awareness. First, we model the optimization scenario and transform the problem-solving into a Markov Decision Process (MDP). Next, we propose R-DQN, a robust DQN (Deep Q-Network) algorithm tailored for heterogeneous tasks and QoS-aware UAV relay communication optimization scenarios. R-DQN introduces corresponding mechanisms in the training process, network structure, and sampling method to improve the effective exploration capability of DQN, making it more robust and suitable for the dynamic constrained optimization scenario tackled. Simulations and experimental results show that the proposed R-DQN algorithm has better convergence and global optimization abilities than other algorithms, such as Dueling DQN, Noisy DQN, and DDQN.

Anomaly detection for multivariate time series is an essential task in the modern industrial field. Although several methods have been developed for anomaly detection, they usually fail to effectively exploit the metrical-temporal correlation and the other dependencies among multiple variables. To address this problem, we propose a stacked attention autoencoder for anomaly detection in multivariate time series (SA2E-AD); it focuses on fully utilizing the metrical and temporal relationships among multivariate time series. We design a multiattention block, alternately containing the temporal attention and metrical attention components in a hierarchical structure to better reconstruct normal time series, which is helpful in distinguishing the anomalies from the normal time series. Meanwhile, a two-stage training strategy is designed to further separate the anomalies from the normal data. Experiments on three publicly available datasets show that SA2E-AD outperforms the advanced baseline methods in detection performance and demonstrate the effectiveness of each part of the process in our method.

In this article, we propose a novel historical snapshot-based ensemble tracker (HSET) to address visual object tracking. Specifically, our HSET tracker collects multiple historical tracker snapshots to model various appearance patterns of the target object during tracking, and performs ensemble operations based on these tracker snapshots to successfully detect the target object. To obtain diverse and representative tracker snapshots for ensemble tracking, we design a tracker snapshot verification scheme to handle dynamical appearance variations of the target object and alleviate unreliable tracker snapshots. Furthermore, the weights of different tracker snapshots are given by an online weight assign algorithm with consideration of both historical appearance information and recent appearance information of the target object. By employing ensemble learning and historical tracker snapshots, the proposed HSET method can get impressive generalization power and tracking robustness to handle significant appearance changes and model drift. Extensive experimental results on public tracking benchmarks indicate that the proposed HSET tracking algorithm reaches encouraging tracking performance compared to multiple state-of-the-art tracking algorithms.

Point-based 3D detection approaches usually suffer from the severe point sampling imbalance problem between foreground and background. We observe that prior works have attempted to alleviate this imbalance by emphasizing foreground sampling. However, even adequate foreground sampling may be extremely unbalanced between nearby and distant objects, yielding unsatisfactory performance in detecting distant objects. To tackle this issue, this paper first proposes a novel method named Distant Object Augmented Set Abstraction and Regression (DO-SA&R) to enhance distant object detection, which is vital for the timely response of decision-making systems like autonomous driving. Technically, our approach first designs DO-SA with novel distant object augmented farthest point sampling (DO-FPS) to emphasize sampling on distant objects by leveraging both object-dependent and depth-dependent information. Then, we propose distant object augmented regression to reweight all the instance boxes for strengthening regression training on distant objects. In practice, the proposed DO-SA&R can be easily embedded into the existing modules, yielding consistent performance improvements, especially on detecting distant objects. Extensive experiments are conducted on the popular KITTI, nuScenes and Waymo datasets, and DO-SA&R demonstrates superior performance, especially for distant object detection. Our code is available at https://github.com/mikasa3lili/DO-SAR.

In this article, a novel multitask sparse correlation filters (MTSCF) model, which introduces multitask sparse learning into the CFs framework, is proposed for visual tracking. Specifically, the proposed MTSCF method exploits multitask learning to take the interdependencies among different visual features (e.g., histogram of oriented gradient (HOG), color names, and CNN features) into account to simultaneously learn the CFs and make the learned filters enhance and complement each other to boost the tracking performance. Moreover, it also performs feature selection to dynamically select discriminative spatial features from the target region to distinguish the target object from the background. A$l_{2,1}$regularization term is considered to realize multitask sparse learning. In order to solve the objective model, alternating direction method of multipliers is utilized for learning the CFs. By considering multitask sparse learning, the proposed MTSCF model can fully utilize the strength of different visual features and select effective spatial features to better model the appearance of the target object. Extensive experiment results on multiple tracking benchmarks demonstrate that our MTSCF tracker achieves competitive tracking performance in comparison with several state-of-the-art trackers.

Correlation filters with Convolutional Neural Networks (CNNs) features have obtained tremendous attention and success in visual tracking. However, redundant and noisy feature channels existed in CNN features may cause severe over-fitting and greatly limit the discriminative power of the tracking model. To tackle the issue, in this paper, we develop a new and effective channel-aware correlation filters (CACF) method for boosting the tracking performance. Our CACF method aims to dynamically select representative and discriminative feature channels from high-dimensional CNN features to reduce the model complexity and better distinguish the target object from the background. Moreover, the CACF model is solved by the alternating direction method of multipliers (ADMM) to learn correlation filters. By retaining reliable feature channels, our CACF tracking method can reach better generalization ability and discriminative ability to accurately localize the target object. Comprehensive experiments are conducted on challenging tracking datasets, and the experiment results prove that our CACF method obtains favorable tracking accuracy compared to several popular tracking methods.

Object tracking is a key component of self-driving systems and has important meanings to alleviate traffic accidents. Therefore, it is meaningful to design a high performance and real-time tracker for improving the stability and safety of self-driving systems. In this paper, an effective and efficient feature fusion tracker, which dynamically fuses gradient and color features to model the appearance of the target object, is designed with the correlation filters framework for fast tracking. To be specific, two complementary correlation filters for gradient (e.g. HOG) and color (e.g. ColorNames) features are maintained during tracking, and the proposed feature fusion method adaptively adjusts the weights of them to deal with large appearance changes of the target object in challenging tracking scenes. The weights are decided by the consistency of the final tracking result and the predicted results obtained by two correlation filters. Moreover, a failure detection scheme is designed to alleviate the model drift issue caused by undesirable model updates to improve the tracking accuracy. If a tracking result is identified as a failed case, re-detection operations are performed to accurately localize the target object. The experimental results prove that the proposed tracker can achieve competitive tracking performance and a satisfactory tracking speed of 25.3 FPS in comparison with several state-of-the-art trackers on challenging tracking benchmarks.

In this paper, a novel spatial-aware tracker (SAT), which utilizes the Siamese network and multiple correlation filters, is proposed to deal with fast motion and model drift problem in visual tracking. Specifically, the Siamese network is first used by an adaptive spatial search strategy to detect the target object in larger search areas. An extended search patch is generated if the target position obtained by the Siamese network is far away from the previous target position. Then, multiple correlation filters perform detection operations on both the extended search patch and the original search patch. With the proposed spatial selection scheme, SAT can accurately track the target object in challenging tracking scenes. By taking advantage of the Siamese network and multiple correlation filters, the proposed SAT tracker can effectively deal with fast motion and model drift problems to achieve better tracking performance. Extensive experimental results demonstrate that the proposed SAT tracker performs superiorly against several state-of-the-art trackers on OTB-2015 tracking benchmark.

In this paper, we propose a novel local sparse representation-based tracking framework for visual tracking. To deeply mine the appearance characteristics of different local patches, the proposed method divides all local patches of a candidate target into three categories, which are stable patches, valid patches, and invalid patches. All these patches are assigned different weights to consider the different importance of the local patches. For stable patches, we introduce a local sparse score to identify them, and discriminative local sparse coding is developed to decrease the weights of background patches among the stable patches. For valid patches and invalid patches, we adopt local linear regression to distinguish the former from the latter. Furthermore, we propose a weight shrinkage method to determine weights for different valid patches to make our patch weight computation more reasonable. Experimental results on public tracking benchmarks with challenging sequences demonstrate that the proposed method performs favorably against other state-of-the-art tracking methods.