Target search for moving and invisible objects has always been considered a challenge, as the floating objects drift with the flows. This study focuses on target search by multiple autonomous ...underwater vehicles (AUV) and investigates a multi-agent target search method (MATSMI) for moving and invisible objects. In the MATSMI algorithm, based on the multi-agent deep deterministic policy gradient (MADDPG) method, we add spatial and temporal information to the reinforcement learning state and set up specialized rewards in conjunction with a maritime target search scenario. Additionally, we construct a simulation environment to simulate a multi-AUV search for the floating object. The simulation results show that the MATSMI method has about 20% higher search success rate and about 70 steps shorter search time than the traditional search method. In addition, the MATSMI method converges faster than the MADDPG method. This paper provides a novel and effective method for solving the maritime target search problem.
The underwater environment is still unknown for humans, so the high definition camera is an important tool for data acquisition at short distances underwater. Due to insufficient power, the image ...data collected by underwater submersible devices cannot be analyzed in real time. Based on the characteristics of Field-Programmable Gate Array (FPGA), low power consumption, strong computing capability, and high flexibility, we design an embedded FPGA image recognition system on Convolutional Neural Network (CNN). By using two technologies of FPGA, parallelism and pipeline, the parallelization of multi-depth convolution operations is realized. In the experimental phase, we collect and segment the images from underwater video recorded by the submersible. Next, we join the tags with the images to build the training set. The test results show that the proposed FPGA system achieves the same accuracy as the workstation, and we get a frame rate at 25 FPS with the resolution of 1920 × 1080. This meets our needs for underwater identification tasks.
The marine environment presents a unique set of challenges for human-robot interaction. Communicating with gestures is a common way for interacting between the diver and autonomous underwater ...vehicles (AUVs). However, underwater gesture recognition is a challenging visual task for AUVs due to light refraction and wavelength color attenuation issues. Current gesture recognition methods classify the whole image directly or locate the hand position first and then classify the hand features. Among these purely visual approaches, textual information is largely ignored. This paper proposes a visual-textual model for underwater hand gesture recognition (VT-UHGR). The VT-UHGR model encodes the underwater diver's image as visual features, the category text as textual features, and generates visual-textual features through multimodal interactions. We guide AUVs to use image-text matching for learning and inference. The proposed method achieves better performance than most existing purely visual methods on the dataset CADDY, demonstrating the effectiveness of using textual patterns for underwater gesture recognition.
For the interaction between marine robots and divers in the underwater environment, a method of diver’s gesture recognition and segmentation is proposed. This method first uses the progressive ...growing training method to optimize the generative adversarial networks, generating high-resolution images with complex content. Then, we use the generative adversarial network model as a data augmentation method and generate high-resolution images. We make the masks of gestures in the new dataset and use the mask R-CNN algorithm for gesture recognition and gesture segmentation. The experimental results show that the generating data improves the accuracy of several object recognition algorithms but cannot completely replace the original data and the mean average precision of gesture recognition is 0.85. The visualization shows the validity and weakness of segmentation.
Working in collaboration with an Autonomous Underwater Vehicle is a new working method for divers. Using gestures to give instructions for the diver is a simple and effective mode of underwater ...human–robot interaction (HRI). In this paper, a gestures tracking method for under human–robot interaction based on fuzzy control is proposed. Firstly, four object recognition algorithms in terms of gesture recognition are compared. YOLO V4-tiny was an extremely high performance, as the gesture area recognition algorithm. We propose a model based on Siamese Network for gesture classification. A gesture tracking method based on fuzzy control is proposed, analyzing the image from AUV front camera to establish a 3D fuzzy rule set. This method can realize the self-regulation of AUV and keep the diver's gestures in the camera view. The experiment result shows the efficiency of the proposed method.
Skeleton-based action recognition methods are limited by the semantic extraction of spatio-temporal skeletal maps. However, current methods have difficulty in effectively combining features from both ...temporal and spatial graph dimensions and tend to be thick on one side and thin on the other. In this paper, we propose a Temporal-Channel Aggregation Graph Convolutional Networks (TCA-GCN) to learn spatial and temporal topologies dynamically and efficiently aggregate topological features in different temporal and channel dimensions for skeleton-based action recognition. We use the Temporal Aggregation module to learn temporal dimensional features and the Channel Aggregation module to efficiently combine spatial dynamic channel-wise topological features with temporal dynamic topological features. In addition, we extract multi-scale skeletal features on temporal modeling and fuse them with an attention mechanism. Extensive experiments show that our model results outperform state-of-the-art methods on the NTU RGB+D, NTU RGB+D 120, and NW-UCLA datasets.
Tectonically emplaced peridotites from North Hebei Province, North China Craton, have retained an original harzburgite mineral assemblage of olivine （54%-58%） ＋ orthopyroxene （40%-46%） ＋minor ...clinopyroxene （〈1%）＋spinel. Samples with honinite-like chemical compositions also coexist with these peridotites. The spinels within the peridotites have high-A1 end-members with A1203 content of 30 wt%-50 wt%, typical of mantle spinels. When compared with experimentally determined melt extraction trajectories, the harzburgites display a high degree of melting and enrichment of SiO2, which is typical of cratonic mantle peridotites. The peridotites display variably enriched light rare earth elements （REEs）, relatively depleted middle REEs and weakly fractionated heavy REEs, which suggest a melt extraction of over 25% in the spinel stability field. The occurrence of are- and SSZ-type chromian spinels in the peridotites suggests that melt extraction and metasomatism occurred mostly in a subduction-related setting. This is also supported by the geochemical data of the coexisting boninite-like samples. The peridotites have lS7Os/lSSOs ratios ranging from 0.113-0.122, which is typical of cratonic iithospheric mantle. These lSTOs/ISSOs ratios yield model melt extraction ages （TRD） ranging from 981 Ma to 2054 Ma, which may represent the minimum estimation of the melt extraction age. The Ai203- lSTOs/lSSOs-proxy isochron ages of 2.4 Ga-2.7 Ga suggest a mantle melt depletion age between the Late Achaean and Early Paleoproterozoic. Both the peridotites and boninite-like rocks are therefore interpreted as tectonically exhumed continental lithospheric mantle of the North China Craton, which has experienced mantle melt depletion and subduction-related mantle metasomatism during the Neoarchean- Paleoproterozoic.
Rat is a valuable model for pharmacological and physiological studies. Germline-competent rat embryonic stem (rES) cell lines have been successfully established and the molecular networks maintaining ...the self-renewing, undifferentiated state of rES cells have also been well uncovered. However, little is known about the differentiation strategies and the underlying mechanisms of how these authentic rat pluripotent stem cells give rise to specific cell types. The aim of this study is to investigate the neural differentiation capacity of rES cells. By means of a modified procedure based on previous publications – combination of mitogen-activated protein kinase (MAPK) and glycogen synthase kinase 3 (GSK3) inhibitors (two inhibitors, “2i”) with feeder-conditioned medium, we successfully obtained high-quality rat embryoid bodies (rEBs) from rES cells and then differentiated them to tripotent neural progenitors. These rES cell-derived neural progenitor cells (rNPCs) were capable of self-renewing and giving rise to all three neural lineages, including astrocytes, oligodendrocytes, and neurons. Besides, these rES cell-derived neurons stained positive for γ-aminobutyric acid (GABA) and tyrosine hydroxylase (TH). In summary, we develop an experimental system for differentiating rES cells to tripotent neural progenitors, which may provide a powerful tool for pharmacological test and a valuable platform for studying the pathogenesis of many neurodegenerative disorders such as Parkinson's disease and the development of rat nervous system.
PDF
