Driven by deep learning techniques, perception technology in autonomous driving has developed rapidly in recent years, enabling vehicles to accurately detect and interpret surrounding environment for safe and efficient navigation. To achieve accurate and robust perception capabilities, autonomous vehicles are often equipped with multiple sensors, making sensor fusion a crucial part of the perception system. Among these fused sensors, radars and cameras enable a complementary and cost-effective perception of the surrounding environment regardless of lighting and weather conditions. This review aims to provide a comprehensive guideline for radar-camera fusion, particularly concentrating on perception tasks related to object detection and semantic segmentation. Based on the principles of the radar and camera sensors, we delve into the data processing process and representations, followed by an in-depth analysis and summary of radar-camera fusion datasets. In the review of methodologies in radar-camera fusion, we address interrogative questions, including “why to fuse”, “what to fuse”, “where to fuse”, “when to fuse”, and “how to fuse”, subsequently discussing various challenges and potential research directions within this domain. To ease the retrieval and comparison of datasets and fusion methods, we also provide an interactive website: https://radar-camera-fusion.github.io .

Autonomous driving on water surfaces plays an essential role in executing hazardous and time-consuming missions, such as maritime surveillance, survivor rescue, environmental monitoring, hydrography mapping and waste cleaning. This work presents WaterScenes, the first multi-task 4D radar-camera fusion dataset for autonomous driving on water surfaces. Equipped with a 4D radar and a monocular camera, our Unmanned Surface Vehicle (USV) proffers all-weather solutions for discerning object-related information, including color, shape, texture, range, velocity, azimuth, and elevation. Focusing on typical static and dynamic objects on water surfaces, we label the camera images and radar point clouds at pixel-level and point-level, respectively. In addition to basic perception tasks, such as object detection, instance segmentation and semantic segmentation, we also provide annotations for free-space segmentation and waterline segmentation. Leveraging the multi-task and multi-modal data, we conduct benchmark experiments on the uni-modality of radar and camera, as well as the fused modalities. Experimental results demonstrate that 4D radar-camera fusion can considerably improve the accuracy and robustness of perception on water surfaces, especially in adverse lighting and weather conditions. WaterScenes dataset is public on https://waterscenes.github.io.

With the rapid advancements of sensor technology and deep learning, autonomous driving systems are providing safe and efficient access to intelligent vehicles as well as intelligent transportation. Among these equipped sensors, the radar sensor plays a crucial role in providing robust perception information in diverse environmental conditions. This review focuses on exploring different radar data representations utilized in autonomous driving systems. Firstly, we introduce the capabilities and limitations of the radar sensor by examining the working principles of radar perception and signal processing of radar measurements. Then, we delve into the generation process of five radar representations, including the ADC signal, radar tensor, point cloud, grid map, and micro-Doppler signature. For each radar representation, we examine the related datasets, methods, advantages and limitations. Furthermore, we discuss the challenges faced in these data representations and propose potential research directions. Above all, this comprehensive review offers an in-depth insight into how these representations enhance autonomous system capabilities, providing guidance for radar perception researchers. To facilitate retrieval and comparison of different data representations, datasets and methods, we provide an interactive website at https://radar-camera-fusion.github.io/radar.

Path planning and collision avoidance during autonomous navigation in unknown environments is a crucial issue for unmanned surface vehicles (USVs). This paper improves the traditional D* Lite algorithm and achieves multi-goal path planning and collision avoidance for USVs in unknown and complex environments. By expanding the adjacent search range and setting a safe distance for USVs, we solve the issue of limited steering maneuverability in USVs with fewer DOF during autonomous navigation. We propose an approach to optimize the planned path during navigation by comparing the estimated distance with the actual distance between the current waypoint and the goal waypoint. A minimum binary heap is used to optimize the priority queue of the D* Lite and significantly reduce the path search time. Simulation results show that the improved D * Lite can significantly reduce the path planning time, optimize the planned path and solve the issue of limited steering maneuverability in USVs. We apply the algorithm to a real USV for further tests. Experimental results show that the USV can plan an optimized path while avoiding both static and dynamic obstacles in complex environments with a safe distance during autonomous navigation.

[Background] Virtual reality (VR) technology can provide unique immersive experiences for group users, and especially for analytics tasks with visual information in learning. Providing a shared control/view may improve the task performance and enhance the user experience during VR collaboration. [Objectives] Therefore, this research explores the effect of collaborative modes and user position arrangements on task performance, user engagement, and collaboration behaviors and patterns in a VR learning environment that supports immersive collaborative tasks. [Method] The study involved two collaborative modes (shared and non-shared view and control) and three position arrangements (side-by-side, corner-to-corner, and back-to-back). A user study was conducted with 30 participants divided into three groups (Single, Shared, and Non-Shared) using a VR application that allowed users to explore the structural and transformational properties of 3D geometric shapes. [Results] The results showed that the shared mode would lead to higher task performance than single users for learning analytics tasks in VR. Besides, the side-by-side position got a higher score and more favor for enhancing the collaborative experience. [Conclusion] The shared view would be more suitable for improving task performance in collaborative VR. In addition, the side-by-side position may provide a higher user experience when collaborating in learning VR. From these results, a set of guidelines for the design of collaborative visualizations for VR environments are distilled and presented at the end of the paper. All in all, although our experiment is based on a colocated setting with two users, the results are applicable to both colocated and distributed collaborative scenarios with two or more users.

Target selection in dense virtual reality (VR) environments is challenging. Prior work has explored different controller-based raycasting techniques to assist target selection in such environments. However, limited research has focused on selection via mid-air barehand, which represents another major input metaphor for immersive environments. In this paper, we first review the existing raycasting selection techniques for dense VR environments. Based on this, we propose and develop two freehand pointing selection techniques—HandDepthCursor and HandConeGrid, and implement MultiFingerBubble, a recently-proposed technique. We then conduct a user study to compare and evaluate their performance and experience in a target selection task in dense VR environments. Our results suggest that HandDepthCursor and HandConeGrid led to significantly faster and more accurate selection performance, and lower perceived workload and arm fatigue. In addition, HandConeGrid showed a distinct advantage in high-density environments.