Multi-modal tracking gains attention due to its ability be more accurate and robust in complex scenarios compared traditional RGB-based tracking. Its key lies how fuse multi-modal data reduce the gap between modalities. However, still severely suffers from deficiency, thus resulting insufficient learning of fusion modules. Instead building such a module, this paper, we provide new perspective on by attaching importance visual prompts. We design novel prompt tracker (ProTrack), which can...

In the complex and stochastic traffic flow, ensuring safe driving requires improvements in perception decision-making. This paper proposed a decision-control method that leveraged scene understanding capabilities of semantic segmentation networks stable convergence strategies Deep Reinforcement Learning (DRL) algorithms to achieve more accurate effective autonomous decision-control. Perception features obtained from cameras sensors equipped with model were used as input for intelligent...

This study introduced a novel non-thermal approach to improve the emulsifying properties of liquid egg yolk by ozonation technology. The best performance was obtained following 20 min ozone treatment. emulsification activity, emulsion stability, creaming index and capacity were significantly (P < 0.05) increased 42.2%, 175.0%, 35.0% 42.4%, respectively. structural changes proteins monitored UV spectroscopy, endogenous fluorescence circular dichroism zeta potential investigate formation...

Purpose The goal of this research is to develop a dynamic step path planning algorithm based on the rapidly exploring random tree (RRT) that combines Q-learning with Gaussian distribution obstacles. A route for autonomous vehicles may be swiftly created using algorithm. Design/methodology/approach issue divided into three key steps by authors. First, expansion sped up size combination and invalid nodes are then removed from initially pathways bidirectional pruning. B-splines employed smooth...

Given its importance, fault diagnosis has attracted considerable attention in the literature, and several machine learning methods have been proposed to discover characteristics of different aspects diagnosis. In this paper, we propose a Hybrid Deep Belief Network (HDBN) model that integrates data ways for intelligent motor drive systems, such as vehicle system. particular, three fusion methods: union, join, hybrid, based on detailed research. Additionally, significance is explained from...

Thanks to the development of Internet things (IoT) and edge computing, smart cameras across cities provide a massive amount image samples with time location labels, laying solid basis for deep mining information in-depth decision analysis. Therefore, this paper proposes convert images spatiotemporal labels into quantifiable data on emotions, apply them crime prediction. Firstly, human emotions were divided three categories: negative, neutral, positive. Then, facial expression recognition...

Articulated robot is now widely employed in manufacture, such as, welding, painting, and assembly, with high precision endurance. It plays an important role scientific technological innovation. Trajectory planning of articulated one the key researches industrial robot. The commonly used trajectory algorithm robot, polynomial interpolation joint space linear Cartesian are introduced. Researches on surveyed. Meanwhile these algorithms analysed. Some further developing trend indicated.

This paper aims to analyze the intrinsic characteristics of urban crime in China by quantifying data original case record. By comparing predicted result situation with observation, characteristic and its law are validated. Firstly, a quantitative method information based on Chinese description is developed, which can be used transform unstructured record security degree model. Secondly, analysis internal case, from number cases, occurrence time location. Thirdly, an autoregressive integrated...

In view of the future lack human resources due to aging population, automatic, Intelligent Mechatronic Systems (IMSs) and Transportation (ITSs) have broad application prospects. However, complex scenarios limited open design make designing highly efficient ITS systems still a challenging task. this paper, optimal load factor solving solution is established. By converting three user requirements including working distance, time into load-related factors, result can be obtained among system...

Crime poses a major threat to human life and property, which has been recognized as one of the most crucial problems in our society. Predicting number crime incidents each region city before they happen is great importance fight against crime. There deal research focused on prediction, ranging from introducing diversified data sources exploring various prediction models. However, existing approaches fail offer fine-scale results take little notice intricate spatial-temporal-categorical...

This study investigates the relationship and genetic mechanisms of liver heart diseases, focusing on liver-heart axis (LHA) as a fundamental biological basis. Through genome-wide association analysis, we explore shared genes pathways related to LHA. Shared factors are found in 8 out 20 pairs, indicating correlations. The analysis reveals 53 loci with pleiotropic effects, including exhibiting causality across multiple traits. Based SNP-p level tissue-specific multi-marker genomic annotation...

With the rapid development of artificial intelligence in recent years, intelligent evaluation college students’ growth by means monitoring data from training processes is becoming a promising technique field education. Current studies, however, tend to utilize course grades, which are objective, predict grade-point averages (GPAs), but usually neglect subjective factors like psychological resilience. To solve this problem, paper takes mechanical engineering as research object, and proposes...

The communication links of multi-agent systems (MASs) generate random time delays, which significantly affect accuracy and real-time control. A Markov chain is established at the input end output a link to express this delay introduced asynchronous data fusion an intelligent location. influence delays on positioning performance investigated. distributed simulation system based MASs employed simulate in multi-sensor data. Different algorithms are tested by changing upper boundary delay. test...

Abstract In this paper, the trajectory tracking problem of position and yaw angle an underactuated quadrotor UAV is studied. Considering that often encounters unknown disturbance resulted by model parameter perturbations environmental changes, a new L 2 ‐gain robust control method presented based on dissipation theory. First, dynamic with disturbances transformed introducing three virtual variables. way, can be implemented dual loop control, is, attitude loop. Then, inequality, controller...

When studying the relationship and potential association between various data, researchers face multiple challenges: first, need to collect data from knowledge systems channels, which challenges breadth of expertise researchers; second, it is challenging perceive changes application scenarios environment through these such as image unstructured that may contain useful information; third, short term behavior prediction with more frequent challenging, especially when sensitivity model...

This paper discusses the problem of robust controller design for two-dimensional (2-D) Markovian jump linear systems. The is demonstrated using Fornasini-Marchesini local state-space models, which are affected by uncertainties. transition-mode probability matrix homogenous and known. It assumed that mode information available implementation. Then, a mode-dependent state-feedback proposed. By substituting into 2-D system, stochastic closed-loop system obtained, because variable, external...

The network teaching system in higher education is playing an increasingly important role modernization construction.Through the teaching, creating a digital learning environment, to promote reform of concept, content and method, improve quality students' ability survival development information society.The research this article provide technical support for management development.Based on SQL Server database structure, conceptual structure design, logic design index design.In paper, with...

A new task scheduling algorithm based on Hadoop is proposed to optimize of resources problems under the Distributed cloud-computing platforms.The core idea full reference current network conditions and treat it as an important for system scheduling, with bandwidth management ability ,SDN architecture allows us allocate according a time slot strategy, then operation completed sooner or later decide whether assigned local node low load non-local .In this way, we will not only ensure locality...

In networked control systems, the communication load is a main concern for implementing model predictive (MPC). This paper introduces self-triggered MPC algorithm under network environments to reduce load, discrete-time linear systems with bounded state and output disturbances, when only system can be measured at triggering instants. The proposed mainly based on estimator whose estimation error by an invariant set, of nominal system. Moreover, new cost function explicitly adopts in...

Fisheye imaging technique plays an important role in various areas. However, fisheye lenses can cause severe image distortion, which requires distortion rectification for these images. after rectification, the images will suffer from problem of resolution reduction, and degree reduction varies different regions. This directly affect subsequent process content analysis. Therefore, it is needed to adopt a super-resolution reconstruction algorithm enhance rectified image. paper designs...