Autonomous driving has gradually become a research hotspot in recent years. Visual Simultaneous Localization and Mapping (SLAM) technology can help unmanned vehicles accurately explore the environment at a lower cost, and the readability of the map can be improved by integrating target detection algorithms. However, the location and 3D shape of the object in the map were not obtained. The method of RGBD-SLAM based on object detection with two-stream YOLOv4-MobileNetv3 convolutional neural network is proposed in this paper. RGBD SLAM algorithm and target detection algorithm are combined to build an algorithm model that can generate the global sparse map and build target dense map quickly. The two-stream network is integrated to obtain 2D information about the target, and further combined with the camera pose after the front-end key frame detection of the SLAM algorithm in this paper, and the dense 3D point cloud of the target and the center point position of the object is obtained. Then, the sparse point cloud of the SLAM system and the dense point cloud of the target can be obtained. The experimental results show that the number of point clouds decreases by about 50% and the time for mapping accounts for about 60% of the global dense mapping time. The method of this paper can efficiently decrease the computational space and improve the speed of semantic mapping, which verifies its feasibility and superiority. It can be used to achieve large-area mapping and the ability to update maps during autonomous driving.

With the rapid development of artificial intelligence, a neural network is widely used in various fields. The target detection algorithm is mainly based on a neural network, but the accuracy of the target detection algorithm is greatly related to the complexity of the scene and texture. A target detection algorithm based on the RGB-D image from the perspective of the lightweight of the target detection network model and the integration of the depth map to overcome the weak environmental illumination with self-powered sensors information is proposed. This article analyzes the networkmodel structure of YOLOv4 andMobileNet, compares the variation of parameter numbers between depthwise separable convolution and convolutional neural networks, and combines the advantages of the YOLOv4 network and the MobileNetv3 network. The main network of three effective feature layers in YOLOv4 is replaced by the MobileNetv3 network for initial feature layer extraction to strengthen the feature extraction network. At the same time, the standard convolutionmodels in the network are replaced by depthwise separable convolution. The proposed method is compared with YOLOv4 and YOLOv4-MobileNetv3 in this article, and the experimental results show that the proposed network retains its original advantages in accuracy, but the size of the network model is about 23% of that of the YOLOv4 network model, the processing speed is about 42% higher than that of the YOLOv4 network model, and the detection accuracy can still reach 83% in the environment with poor lighting conditions.

Antibiotic resistance and pathogenic bacterial monitoring are two main challenges in clinical diagnostics and treatment of bacteria. To conquer those issues, many new materials and novel technologies are explored. Among them, quantum dots (QDs) are regarded as not only powerful bacteriostatic agents, but also bacterial monitoring probes due to their excellent photoluminescence properties, tunable multiwavelength luminescence, and capability in reactive oxygen species-generation. While, till now, no systematic and critical review has been published in this related research area. This review paper aims at summarization and review of the monitoring and combating of pathogenic bacteria utilizing QDs (mainly in the recent five years), in view of their classification, modification, mechanism, and application. Semiconductor QDs, transmission-metal-based QDs, carbon-based QDs, and silicon-based QDs are mainly covered in this review. Moreover, a thorough discussion on the possible interaction mechanism between bacteria and QDs is given, which will contribute to the designing of QDs for novel and high-efficient bacterial monitoring and combating applications. Eventually, the current challenges and the future research focus of QDs in bacterial monitoring and combating are proposed. This review will be of significance for the related field to understand the designing principles of QDs for pathogenic bacterial monitoring and combating applications.

SummaryTarget localization in unknown environment is one of the development directions of mobile robots. Simultaneous localization and mapping (SLAM) can be used to build maps in unknown environments, but it has the problem of poor readability and interactivity. In this article, target detection and SLAM are combined to search and locate the target by using rich RGBD images information. The determined position in the global map is conducive to the follow‐up operation of the target by mobile robots. By establishing a local dense point cloud map of the target object, the current state of the target object is directly displayed, the readability of the map is improved, and the disadvantages of difficult understanding of the global sparse map and slow construction of the global dense map are avoided. A target localization algorithm under the framework of yolov4 is designed to apply in the process of SLAM global mapping. Our works are helpful for obtaining positions of objects in three‐dimensional space. The experimental results show that the time‐consuming of this method in dense mapping is reduced by 50%–70%, and the number of point clouds is also reduced by 60%–70%.

Robots are applied more and more widely to realise the cold rolling warehouse area automatic unmanned operation. The application method of Internet of Things (IoT) is adopted to study the key technologies such as the intelligence and automation of the crane, and unmanned transformation is carried out on the basis of manual operation. Through the internet information exchange, crane automatic control, anti-sway control, spreader intelligent transformation and other technical means, to realise the crane equipment automatic operation. The practice results show that, under the application of IoT, the automatic transformation of crane has greatly improved the production level, reduced the production cost, improved the economic benefit, reduced the management cost and reduced the safety hidden danger and provided the theoretical basis and solutions for realising the intelligent operation in the factory warehouse area. Owing the great difference between the actual operating environment and the ideal operating environment, it is necessary to further improve the operating accuracy and system stability, and reduce the error of the crane actuator, so as to realise the omni-directional automatic control of the unmanned crane.

Gesture recognition is one of the important ways of human-computer interaction, which is mainly detected by visual technology. The temporal and spatial features are extracted by convolution of the video containing gesture. However, compared with the convolution calculation of a single image, multiframe image of dynamic gestures has more computation, more complex feature extraction, and more network parameters, which affects the recognition efficiency and real-time performance of the model. To solve above problems, a dynamic gesture recognition model based on CBAM-C3D is proposed. Key frame extraction technology, multimodal joint training, and network optimization with BN layer are used for making the network performance better. The experiments show that the recognition accuracy of the proposed 3D convolutional neural network combined with attention mechanism reaches 72.4% on EgoGesture dataset, which is improved greatly compared with the current main dynamic gesture recognition methods, and the effectiveness of the proposed algorithm is verified.

Detection and localisation of pipeline leakage is an effective measure to save water resources. Generalised cross-correlation algorithm is widely used in leak detection. In this paper, by comparing the weighted algorithms of generalised cross-correlation functions, the GCC-PHAT algorithm is used to design a leak detection and positioning system, and relevant experiments are carried out. Multiple sensors are used (including flowmeters, pressure gauges and vibration sensors) to detect leaks and obtain relevant data in the experiment and locate the leak point through the GCC-PHAT algorithm. It is found that the method has better positioning performance through experiments.

Detection and localization of pipeline leakage is one of the important means to save water resources. A set of leakage detection was set up and experiments were carried out. The system contains flowmeter, pressure gauges and vibration sensor. The signal was continuous collected to constantly detect and alert water leaks. The leakage location is based on GCC-PHAT algorithm. a leakage detection and location system based on GCC-PHAT algorithm is proposed. Experiments have verified that our system works well.

Reinforcement learning algorithms generally belong to one of two categories: model-based techniques, which attempt to overcome the issue of a lack of prior knowledge by enabling the agent to construct a representation of its environment, and model-free techniques, which learn a direct mapping from states to actions. Model-free approaches are typically less practical because running a simulation is very time consuming or expensive, and model-based approaches tend to achieve lower asymptotic performance due to the error in model approximation. To design an effective model-based algorithm, Janner, M. et al [1]. studies the role of model usage in policy optimization and introduces a practical algorithm called model-based policy optimization (MBPO), which makes limited use of a predictive model to achieve pronounced improvements in performance comparing to other model-based approaches. The authors of the paper [1] first formulate and analyze a general implementation for MBPO with monotonic improvement at each step which uses a predictive model to optimize policy and utilizes the policy to collect data and train the model. Previous study shows it is difficult to justify model usage due to pessimistic bounds on model error and this paper finds a way to modify the pessimistic bounds which solves the problem. Based on the analysis, the authors propose a simple model-based reinforcement learning algorithm of using short model-generated rollouts branched from real data to improve model effectiveness. The experiments show this algorithm is faster than other state-ofthe-art model-based methods such as STEVE [2], and matches the performance of the best model-free methods like SAC [3]. In this reproducibility report, we study in detail the MBPO algorithm described in the paper (detailed in Section 3). Our work mainly focuses on the replication of their algorithm and the re-implementation of their predictive model but in a PyTorch version (detailed in Section 4). Lastly, we describe our experiments and provide insightful analysis of our results (detailed in Section 5). We provide the source code1 for generating the results and setting up the experiments.