Due to the explosive increase in online videos, near-duplicate video retrieval (NDVR) has attracted much researcher attention. NDVR very wide applications, such as copyright protection, monitoring, and automatic tagging. Local features serve elementary building blocks many algorithms, most of them exploit local volume information using a bag (BOF) representation. However, representation ignores potentially valuable about global distribution interest points. Moreover, discriminative power...

As a fundamental element of the transportation system, traffic signs are widely used to guide behaviors. In recent years, drones have emerged as an important tool for monitoring conditions signs. However, existing image processing technique is heavily reliant on annotations. It time consuming build high-quality dataset with diverse training images and human this paper, we introduce utilization Vision–language Models (VLMs) in sign detection task. Without need discrete labels, rapid...

This paper presents an in-depth study and analysis of the image feature extraction technique for ancient ceramic identification using algorithm partial differential equations. Image features ceramics are closely related to specific raw material selection process technology, complete acquisition is a prerequisite achieving ceramics, since quality extracted area-grown method background pixels does not have generalizability. In this paper, we propose deep learning-based method, Eased as...

A key instrument for upgrading China’s agriculture is the Internet of Things (IoT). To solve problem IoT technology promotion, farmers’ intentions to adopt must be transformed into behavior, and their behaviors unified. The multivariate logistic model was used analyses factors influencing intention behavioral deviation based on survey data vegetable farmers in Jiangxi Province. ISM investigate relationship hierarchy between deviation. findings revealed that first, a significant exists...

The standard of a golden course should be characterized by advanced, innovative, and challenging features, with its primary objective being the attainment deep learning. In online-to-offline hybrid teaching, online learning comes first, while offline is re-application expansion knowledge. Therefore, widespread participation segment has an important impact on implementation This paper focuses part blended combined self-regulated theory, gives intervention strategies for teaching. Using web...

As a well-known nonlinear tool, mathematical morphology (MM) is still active in image processing. Benefiting from the fixed structuring element (SE), traditional MM (TMM) gets solid theoretical foundation. However, due to inherent diversity of pixels an image, rigid SE paradigm not always practical. result, development with adaptive SE, known as (AMM), has been significant challenge. In this work, we present novel approach for designing using \boldsymbolα-cut fuzzy set. By implementing...

Abstract Removal of helium ash and the anomalous transport deuterium (D) tritium (T) ions driven by collisionless trapped electron mode (CTEM) turbulence in tokamak plasmas with weak magnetic shear are studied. We derive eigenvalue CTEM ash, calculate quasi-linear turbulent fluxes D T simultaneously. Based on analytical results, parametric dependence instability as well D-T is investigated, order to explore parameter region that favorable for expelling more than ions. It found higher...

Multi-task learning has been widely applied in computational vision, natural language processing and other fields, which achieved well performance. In recent years, a lot of work about multi-task recommender system yielded, but there is no previous literature to summarize these works. To bridge this gap, we provide systematic survey systems, aiming help researchers practitioners quickly understand the current progress direction. survey, first introduce background motivation learning-based...

Anomaly detection algorithms (ADA) have been widely used as services in many maintenance monitoring platforms. However, there are numerous that could be applied to these fast changing stream data. Furthermore, IoT data due its dynamic nature, the phenomena of conception drift happened. Therefore, it is a challenging task choose suitable anomaly service (ADS) real time. For accurate online anomalous detection, this paper developed selection method select and configure ADS at run-time....

A machine vision system for SMT-mounting usually involves a two-stage algorithm. It first measures the center and rotate angle of SMD, then checks each electronic component. The component positioning is an important research content in completely automatic SMM. In this paper new high-speed algorithm proposed to locate During locating procedures, it demarcates least image area using four limiting checking lines. Then SMD calculated area. After that, precision, speed characteristic are...

Circle detection and ascertaining its parameters are an important task in computer vision pattern recognition. The method Hough transform is usually applied this process, but algorithm too complex, slow needs abundant memory especially when we deal with complicated background image interferential information. According to the requirement of my research problem, a fast simple should be needed. In order overcome paper put forward new detect circle's center radius. Select initial point region...

Nowadays, an Internet of Things (IoT) device consists algorithms, datasets, and models. Due to good performance deep learning methods, many devices integrated well-trained models in them. IoT empowers users communicate control physical achieve vital information. However, these are vulnerable adversarial attacks, which largely bring potential risks the normal application methods. For instance, very little changes even one point time-series data could lead unreliable or wrong decisions....

Abstract The effects of alpha ( α ) particles on the transport helium ash driven by collisionless trapped electron mode (CTEM) turbulence are analytically studied using quasi-linear theory in tokamak deuterium (D) and tritium (T) plasmas. Under parameters used this work, is mainly determined diffusion due to very weak convection. It found that ratio between diffusivity effective thermal conductivity D He / χ eff CTEM turbulence, which a proper normalized parameter for quantifying efficiency...

Many anomaly detection algorithms have been provided as services for more convenient and efficient utilization in IoT era. Due to concept drift existed dynamic stream data, it is a changeling task apply proper at run time. For effective on-line anomalous data discovery, this paper proposes service selection framework dynamically select configure services. A fast classification model based on XGBoost trained identify the pattern of various so that suitable can be selected configured according...

In the cloud computing environment, release and dynamic deployment of microservices face many challenges, services' dependency is an important consideration for service deployment. The call chain logs record rich information, such as time delay each in business tracking process, present form a chain. Existing research does not fully consider local composite dependencies their discontinuous order to obtain these complex provide basic supporting data adjustment microservices, model proposed,...

Using clustering method to detect useful patterns in large datasets has attracted considerable interest recently. The HKM algorithm (Hierarchical K-means) is very efficient large-scale data analysis. It been widely used build visual vocabulary for scale video/image retrieval system. However, the speed and even accuracy of hierarchical K-means still have room be improved. In this paper, we propose a Parallel N-path quantification which improves on following ways. Firstly, replace Euclidean...

Abstract In contemporary society, career readiness holds paramount significance for individual life, exerting a direct influence on initial employment, job satisfaction, and the sense of identity. Framed within multidimensional item response theory text mining, this study embarks exploring assessment methodologies high school students’ by revising “Career Readiness Questionnaire – Adolescent Version” employing mining techniques. Study One collected 1261 valid data points through cluster...

Abstract In the coal mining production process, wire ropes play a crucial role as main tool for transporting personnel and materials in entire mine hoisting system, their fault identification is particularly critical. To address problem of ropes, recognition model design combining K Singular Value Decomposition (SVD) with Genetic Algorithm (GA) optimized Support Vector Machine (SVM) proposed. Firstly, used to denoise collected rope damage signals. Feature vectors are extracted from denoised...

With the rapid growth of internet usage for enterprise-wide and cross-enterprise business applications (such as those in Electronic Commerce), workflow systems are gaining importance an infrastructure automating inter-organizational interactions. However, traditional centralized management technology can no longer meet needs current application services. For example, e-commerce, may cause many security problems cross-domain implementation workflow. At same time, due to uncertainty...

Non-local block (NLB) is a breakthrough technology in computer vision. It greatly boosts the capability of deep convolutional neural networks (CNNs) to capture long-range dependencies. As critical component NLB, non-local operation can be considered network-based implementation well-known means filter (NLM). Drawing on solid theoretical foundation NLM, we provide an innovative interpretation operation. Specifically, it formulated as optimization problem regularized by Shannon entropy with...