While unmanned aerial vehicle (UAV) technology brings convenience to modern life, it also leads some problems. To achieve anti-UAV, the object detection of UAV is key. YOLO v3, one single-stage detectors, has best performance for balancing accuracy and speed through capturing deep high-level features. In basis this paper improves detect more precisely it's first time introduce v3 based algorithm anti-UAV. It adopts last four scales feature maps instead three predict bounding boxes objects,...

Medical image classification is an important task in computer-aided diagnosis systems. Its performance critically determined by the descriptiveness and discriminative power of features extracted from images. With rapid development deep learning, convolutional neural networks (CNNs) have been widely used to learn optimal high-level raw pixels images for a given task. However, due limited amount labeled medical with certain quality distortions, such techniques crucially suffer training...

In the age of Internet Things and Industrial 4.0, new advanced methods need to be proposed analyse massive multi-source heterogeneous data from rotating machinery since traditional analysis are difficult mine features effectively provide accurate fault results automatically. This paper proposes a rotor unbalance diagnosis method using deep belief network (DBN) learn representative automatically accurately identify states. Multi-source information composed with vibration signal shaft orbit...

Land cover classification using polarimetric synthetic aperture radar (PolSAR) images is an important tool for remote sensing analysis. In view that PolSAR image effective interpretation commonly affected by the absence of discriminative features and presence speckle noises, this letter proposes adaptive nonlocal stacked sparse autoencoder (ANSSAE) to achieve classification. It extracts spatial information adaptively calculating weighted average values each pixel from regions, which can...

Marine floating raft aquaculture (MFRA) is a primary human activity on oceanic resources utilization and development, the physical structure of which so particular that conventional optical remote sensing images cannot discover thoroughly. GaoFen-3 (GF-3), as first launched full polarimetric synthetic aperture radar satellite China, can offer unique opportunity to perform long-term precise monitoring nearshore MFRA. In this paper, GF-3 product description discussed choose applicable modes...

Older adults were disproportionately affected by COVID-19. The purpose of this study was to explore experiences sudden-onset social isolation and factors that influenced it among in two groups older adults. A qualitative thematic with a survey component conducted comparing 18 groups: 12 reporting physical health challenges 6 no challenges. Three themes describe (a) avoiding risk personal as the reason postpone healthcare, (b) grieving church friends lost connection, (c) compound stress due...

Bearing fault diagnosis is of great significance to ensure the safe operation mechanical equipment. This paper proposes an intelligent method rolling bearings based on deep belief network (DBN) with hyperparameter optimization by using parallel computing. Different traditional methods that extract features manually depending much prior knowledge about signal processing techniques and diagnostic expertise, DBN extracts automatically machine learning mechanism. Considering time consuming...

This paper proposes the adoption of a deep learning method to automate categorisation online discussion messages according phases cognitive presence, fundamental construct from widely used Community Inquiry (CoI) framework learning. We investigated not only performance classifier but also its generalisability and interpretability, using explainable artificial intelligence algorithms. In study, we compared Convolution Neural Network (CNN) model with previous approaches reported on literature...

Floating raft aquaculture is a significant part of the mariculture, and its recognition conducive to dynamic monitoring sea area. The satellite synthetic aperture radar (SAR) image can overcome uncertainties marine meteorological environment reflect location floating aquaculture. However, oceanic SAR images are seriously contaminated by speckle noise, effective features deficient, which make difficult. Modified statistical region merging (SRM) algorithm utilized accomplish superpixels...

This paper reports on early stages of a machine learning research project, where phases cognitive presence in MOOC discussions were manually coded preparation for training automated classifiers. We present manual-classification rubric combining Garrison, Anderson and Archer's [11] coding scheme with Park's [25] revised version target MOOC. The inter-rater reliability between two raters achieved 95.4% agreement Cohen's weighted kappa 0.96, demonstrating our classification is plausible the...

This paper proposes a network for low-resolution infrared image target detection in the land battlefield based on YOLO v3 network. By reducing number of pooling layers, risk small miss is reduced. In addition, when training samples limited, strategy using visible light to assist model proposed, which effectively reduces over-fitting. Experiments show that proposed has better performance detecting targets images, and can reduce over-fitting during through visible-light image-assisted strategies.

Automatic analysis of the myriad discussion messages in large online courses can support effective educator-learner interaction at scale. Robust classifiers are an essential foundation for use automatic cognitive presence practice. This study reports on application a revised machine learning approach, which was originally developed from traditional, small-scale, for-credit, courses, to automatically identify phases discussions Philosophy Massive Open Online Course (MOOC). The classifier...

The closed working condition of grouting materials leads to insufficient curing water supply during their strength development. In this work, the superabsorbent polymer (SAP) was used as internal agent for material. effects SAP content on rheological and mechanical properties material were studied under various water-cement ratios (w/c) environments. time-dependent variation fresh slurry’s viscosity measured by Brinell viscometer. pore structure micromorphology characteristics analyzed...

Web service constitutes a new computing model for application. The application based on SOA is promising trend of distributed computing. However, currently system, which uses centralized UDDI to publish and discover has become the bottleneck whole system. In addition, description WSDL cannot support discovery at semantic level. Therefore, we propose fully decentralized interoperable method. P2P network used as basis replace traditional UDDI. At same time, take advantage OWL-S add valuable...

As a typical example of the SOA (service oriented architecture), Web service has been applied in e-business, e-government and many other fields as for its encapsulated, loosely coupled, dynamic binding, etc. mechanism. However, mechanism publishing discovery with centralized architecture, such UDDI, restricts reliability scalability environment services resources on web are fast emerging. Using P2P technique is good solution. This paper discusses an architecture publishing, invocation based...