Energy constraints significantly limit the long-term operation of underwater wireless sensor networks (UWSNs) due to their battery-powered sensor nodes (SNs). In this paper, we investigate an autonomous underwater vehicle (AUV)assisted UWSN where the AUV simultaneously collects data and recharges multiple independently located SNs. We employ the value of information (VoI) to evaluate the importance of the sensing data. We propose a novel Lyapunov-guided deep reinforcement learning (LDRL) algorithm to maximize the long-term average VoI while guaranteeing the battery energy constraints of SNs. We first employ a Lyapunov optimization to decompose the multi-stage stochastic VoI maximization problem into a series of single-stage deterministic subproblems. The Lyapunov drift-pluspenalty function is designed to deal with the long-term energy queue equilibrium. Then, we employ a deep neural network (DNN) to optimize the AUV's path and integrate a sequential convex approximation (SCA) optimization module for optimal charging time allocation. Experimental results demonstrate that our algorithm significantly enhances the long-term average VoI while ensuring the SNs' battery energy constraints.

Utilizing an autonomous underwater vehicle (AUV) for data collection in underwater wireless sensor networks (UWSNs) is a promising approach. However, since underwater sensor nodes are typically battery-powered and difficult to replace, effective energy management is crucial for extending the network lifetime of UWSNs. Additionally, the limited energy of the AUV presents further challenges. To address these issues, this paper proposes an energy-aware AUV-assisted data collection scheme based on dynamic clustering and cluster head selection (DCCHS) to maximize network lifetime. Specifically, we utilize the energy center to define the cluster center and employ a bottom-up hierarchical clustering approach to address the node dynamic clustering problem under the AUV movement distance constraint. Subsequently, we introduce a cluster head (CH) selection algorithm based on iterative optimization, and adds auxiliary CHs near the AUV path to reduce the energy consumption of CHs. Simulation results demonstrate that the proposed DCCHS scheme significantly extends the network lifetime compared to existing schemes, particularly in scenarios with dense node deployment.

Reliable and timely data collection is an important and challenging problem for underwater wireless sensor networks (UWSNs), partly due to the very slow underwater communication and the difficulty in recharging the sensors. In this paper, we exploit the use of autonomous underwater vehicles (AUVs) for UWSN data collection task. Specifically, we investigate how to maximize the value of information (VoI) for data collection through joint optimization of cluster head (CH) selection and multi-AUV path planning. We formulate the joint optimization problem for the task, taking into account the energy constraints of sensor nodes. To solve the problem, we propose a deep reinforcement learning algorithm based on an encoder-decoder architecture. The entire UWSN system is fed into the encoder network, followed by a composite decoder consisting of an AUV selection decoder and a cluster access sequence decoder to obtain the cluster access sequence for each AUV. Based on the determined sequences, we further utilize the dynamic programming algorithm to achieve optimal CH selection. Finally, we obtain the sequence of AUVs accessing the selected CHs. Simulation results demonstrate that the proposed learning-based approach converges and achieves a higher VoI than the existing benchmark algorithms.

In autonomous driving, the LiDAR-Camera system plays a crucial role in a vehicle’s perception of 3D environments. To effectively fuse information from both camera and LiDAR, extrinsic calibration is indispensable. Recently, some researchers have proposed deep learning-based methods that utilize convolutional networks to automatically extract features from LiDAR depth images and RGB images for calibration. However, these features do not sufficiently interact during feature matching, which limits the calibration accuracy. To this end, we introduce a novel extrinsic calibration network (HIFM-Net) in this paper. It establishes a comprehensive connection between camera and LiDAR features by calculating a globally-aware map-to-map cost volume and hierachical point-to-map cost volumes. The former is used to regress large extrinsic offsets. The latter is employed to iteratively fine-tune extrinsic parameters, while the rigidity of LiDAR points is considered in each iteration to enhance regression robustness. Extensive experiments on the KITTI-odometry dataset demonstrate the superior performance of our HIFMNet compared to other state-of-the-art learning-based methods.

Point clouds play a crucial role in robots and intelligent vehicles. Noise caused by fog droplets seriously degrades the quality of point clouds. Previous researches have shown that the extent of degradation is correlated with visibility. The fog attenuation coefficient is associated with visibility. In light of this background, this paper proposes a noise-distribution-based defogging method for point clouds. Our approach hinges on the estimation of the fog attenuation coefficient, facilitated by road-based prior knowledge. Subsequently, our method integrates the fog-induced noise distribution inferred from the LiDAR imaging model with the spatially non-uniform distribution of point clouds caused by LiDAR structure. The fused results are input to a statistical filter based on the relative sparsity of noise to achieve defogging. This paper is one of the early works focusing on point cloud defogging. Its core insight lies in the estimation of the attenuation coefficient and the employment of fog-induced noise distribution for defogging. Experiments demonstrate that our method can accurately mitigate the impact of fog and meanwhile enhance the performance of 3D object detection network.

For 3D LIDAR object detection, data augmentation is an important module to make full use of precious annotated data. As a widely used data augmentation method, GT-aug effectively improves detection performance by inserting sampled groundtruths into LIDAR frames. However, they are often placed in unreasonable areas, leading to the loss of the semantic information between targets and backgrounds during training. To address this problem, we propose a context-aware data augmentation method (CA-aug), which ensures the proper placement of inserted objects by a simple strategy and produces realistic augmented scenes. CA-aug is lightweight and compatible with other augmentation methods. Experiments conducted on KITTI benckmark show that compared with the GT-aug and the similar method in LIDAR-aug (SOTA), it brings higher accuracy to the existing models especially for the detection of cyclists and perdestrians. We also present an in-depth study of augmentation strategies for the range-view-based (RV-based) models and demonstrate that CA-aug can fully exploit the potential of RV-based networks, boosting the moderate mAP of our test model by 8%.