Real-time object detection on unmanned aerial vehicles (UAVs) poses a challenging issue due to the limited computing resources of edge devices. To address this problem, we propose novel lightweight network named LWUAVDet for real-time UAV applications. The detector comprises three core components: E-FPN, PixED Head, and Aux Head. Firstly, develop an extended refined topology in Neck layer, called enhance multi-scale representation each stage alleviate aliasing effect caused by repetitive...

The main challenge of scene classification is to understand the semantic context information high-resolution remote sensing images. Although vision transformer (ViT)-based methods have been explored boost long-range dependencies images, connectivity between neighboring windows still limited. Meanwhile, ViT-based commonly contain a large number parameters, resulting in huge computational consumption. In this paper, novel lightweight dual-branch swin (LDBST) method for proposed, and...

Monitoring marine debris has long been a challenging issue owing to the complex and changeable underwater environment. To fast accurately detect debris, in this article, novel object detection network termed as YOLOTrashCan is proposed for detecting debris. The model consists of feature enhancement fusion. In part, ECA_DO-Conv_CSPDarknet53 backbone, which combines efficient channel attention (ECA) module depthwise over-parameterized convolutional (DO-Conv), extract depth semantic features...

In recent decades, remote sensing object counting has attracted increasing attention from academia and industry due to its potential benefits in urban traffic, public safety, road planning. However, this issue is becoming a challenge for computer vision because of various technical barriers, such as large-scale variation, complex background interference, nonuniform density distribution. Recent results show hopeful prospects using convolutional neural networks (CNNs), but most existing...

Violence behavior detection has played an important role in computer vision, its widely used unmanned security monitoring systems, Internet video filtration, etc. However, automatically detecting violence from surveillance cameras long been a challenging issue due to the real-time and accuracy. In this brief, novel multi-scale spatio-temporal network termed as MSTN is proposed detect stream. To begin with, feature extraction module (STM) developed extract key features between foreground...

An emerging Internet of Video Things (IoVT) application, crowd counting is a computer vision task where the number heads in crowded scene estimated. In recent years, it has attracted increasing attention from academia and industry because its great potential value public safety urban planning. However, become challenge to cross gap between increasingly heavy complex network architecture widely used for pursuit with high accuracy constrained computing storage resources edge environment. To...

Knowledge-grounded Dialogue Response Generation (KRG) can facilitate informative and fidelity dialogues using external knowledge. Prior monolingual works only use the knowledge of corresponding native language. Thus, due to prohibitive costs collecting constructing bases, limited scale accessible always constrains ability KRG, especially in low-resource language scenarios. To this end, we propose a new task, Multi-Source Multilingual Knowledge-Grounded (MMKRG), which simultaneously uses...

Version issues are becoming more and prominent with the continuous development of software. Bug localization for version is time-consuming labor-intensive. Although some bug techniques, such as those based on information retrieval (IR), have been proposed, they cannot handle these bugs very well version-related their own defect patterns. However, few existing works focused revealing patterns utilizing them bugs. To fill this gap, we propose a new approach by leveraging to localize integrated...

The prediction of chaotic time series can well describe the physical properties and influencing factors a sequence numerical data points in successive order has played an increasingly crucial role various scientific engineering communities. These require model to have strong dynamic feature representation ability, which often leads excessive calculation burden. As emerging effective learning method, broad system (BLS) be trained quickly by incremental learning, reconstructed without...

When using very high-resolution (VHR) remote sensing images acquired at different times to detect building changes, the positional inconsistencies caused by satellite imaging angles are an outstanding issue. To tackle this problem, a novel change detection method based on pseudo-orthorectification (PO) is proposed. First, determine displacement value, fast line used accurately extract vertical facade contour lines under constraint of Object Space Positioning Consistency (OSPC). Second, roof...

In this paper, a new type of magnetic wheeled wall-climbing robot is proposed in order to overcome the shortcomings existing climbing robots. It takes advantage SOLIDWORKS establish three-dimensional model. Also, paper introduces design idea robot, uses MATLAB do numerical analysis and make simulation through ANSOFT MAXWELL.

Abstract Background Dexmedetomidine has controversial influence on cardiac electrophysiology. The aim of this study was to explore the effects dexmedetomidine perioperative electrophysiology in patients undergoing general anesthesia. Methods Eighty-one were randomly divided into four groups: groups D 1 , 2 3 receiving 1, 0.5 μg/kg over 10 min and 0.5, μg/kg/h continuous infusion respectively, control group (group C) normal saline. Twelve-lead electrocardiograms recorded at time before...

In recent years, digital watermarking is playing an increasingly important role in the geographical information copyright protection. According to characteristics of GIS vector data, watermark embedding algorithms should at least meet following demands: no significant loss data accuracy, obvious reduction quality and change visual observation. this paper, watermarked inspected through polygon closure, topology, error analysis analysis. And above functions are achieved by coding. Through...

The copyright of GIS vector data has become an increasingly complex topic. main technologies used for protection is the marking techniques delegated by watermarking. In this paper, current status and future prospects summarized from basic theory, generating, embedding, detection real-time monitoring.

Limited by the characters of GIS vector data, such as variety and complexity expression, universality, mass, disorder pretty good privacy, etc., traditional copyright protection methods for image audio productions cannot be applied into data directly. In this paper, a solution to is proposed. Firstly, information characteristics, organization storage characteristics attack are comprehensively analyzed. Secondly, based on hash, file filter driver, dynamic encryption decryption, key techniques...

Abstract Background Directly discharging livestock and poultry slaughter blood without proper treatment can cause severe ecological damage. Exploring the use of microorganisms to break down waste into smaller molecules such as peptides amino acids, well investigating possibility transforming these small water-soluble fertilizers containing holds significant research value in comprehensive utilization blood. Results In this study, a single strain Bacillus pumilus NWMCC0302, which has...