With the rapid development of unmanned aerial vehicle (UAV) industry, UAV image object detection technology has become a hotspot. However, due to large number dense small objects in images, quickly and effectively detecting achieving accurate classification is still challenge. this observation, we propose an efficient network for images based on cross-layer feature aggregation (CFANet). Firstly, design novel cross-feature module (CFA) aggregate features at different scales basis avoiding...

Unmanned Aerial Vehicle (UAV) image target detection holds significant value for a wide range of applications in modern society. However, due to the variable flight altitude UAV, captured images often exhibit differences at scale and contain large number small targets. The existing methods are difficult adapt these changes, resulting decrease accuracy. To address this issue, paper proposes new method UAV object based on full feature aggregation grouped reconstruction FFAGRNet. Firstly,...

Object detection on UAV images is a recent research hotspot. Existing object methods have achieved good results general scenes, but there are inherent challenges with images. The accuracy of limited by complex backgrounds, significant scale differences, and densely arranged small objects. To solve these problems, we propose image network based Self-attention Guidance Multi-scale Feature fusion (SGMFNet). Firstly, design Global-Local module (GLFG). This can effectively combine local...

Recently, video saliency prediction has attracted increasing attention, yet the improvement of its accuracy is still subject to insufficient use multi-scale spatiotemporal features. To address this issue, we propose a 3D convolutional Multi-scale Spatiotemporal Feature Fusion Network (MSFFNet) achieve full utilization Specifically, Bi-directional Temporal-Spatial Pyramid (BiTSFP), first application bi-directional fusion architectures in field, which adds flow shallow location information on...

Aiming at solving the trouble that digital image information is easily intercepted and tampered during transmission, we proposed a color encryption scheme based on alternate quantum random walk controlled Rubik's Cube transformation. At first, separated into three channels: channel R, G B. Besides, sequence generated by walk. Then six faces of are decomposed arranged in specific order two-dimensional plane, each pixel randomly mapped to Cube. The whirling realize scrambling encryption....

Perceptual Video Compression (PVC) is a promising approach to enhancing compression efficiency. The Human Visual System (HVS) possesses many important perceptual characteristics, which can be utilized further enhance encoding efficiency without significantly degrading quality. This paper addresses the issue that existing video methods have not fully leveraged HVS characteristics by proposing scheme, SJ-PVC, uses Just Noticeable Distortion (JND) estimation model based on characteristics....

Multiview stereopsis (MVS) methods, which can reconstruct both the 3D geometry and texture from multiple images, have been rapidly developed extensively investigated feature engineering methods to data-driven ones. However, there is no dataset containing of large-scale scenes high-resolution observations small details benchmark algorithms. To this end, we present GigaMVS, first gigapixel-image-based reconstruction for ultra-large-scale scenes. The gigapixel with wide field-of-view details,...

A Real Time Power Monitoring (RTPM) System is proposed in which end-use detailed energy consumption data provided for each load level. The will be integrated with a BIM (Building Information Modeling) Model to create Realtime on-line electronic Model. This paper describes the RTPM system and process leading its creation. basic components of were designed tested using prototype board.

The growing scale and complexity of the enterprise computing systems under distributed heterogeneous environments present new challenges to system development, integration, maintenance. In this paper, we a model driven Web service development framework combat these challenges. capitalizes on UML profile for Enterprise Distributed Object Computing (EDOC), MDA services. Within framework, first, platform independent models (PIMs) are created using EDOC profile. Second, PIMs broken down into sub...

This study used hierarchical linear modeling (HLM) to investigate the differences in mobile game addiction tendencies among universities (i.e., double first-class universities, ordinary colleges and higher vocational colleges), examine influencing factors of tendency at individual university levels. The participants this were 4,024 college students who completed Chinese Mobile Phone Game Addiction Scale, UCLA Loneliness Positive Negative Affect Scale. ( β = 0.052), positive emotions −0.126),...

To accurately detect multi-scale remote sensing objects in complex backgrounds, we propose a novel transformer-based adaptive object detection method. The backbone network of the method is dual attention vision transformer that utilizes spatial window and channel group to capture feature interactions between different scenes. We further design an path aggregation network. In designed network, CBAM (Convolutional Block Attention Module) utilized suppress background information fusion paths...

Fine-grained visual classification aims to recognize images belonging multiple sub-categories within a same category. It is challenging task due the inherently subtle variations among highly-confused categories. Most existing methods only take an individual image as input, which may limit ability of models contrastive clues from different images. In this paper, we propose effective method called progressive co-attention network (PCA-Net) tackle problem. Specifically, calculate channel-wise...

Current video saliency prediction methods have made great progress relying on the feature extraction capability of CNN, but there are still many defects in hierarchical fusion, limiting further improvement accuracy. To address this issue, we propose a 3D convolutional Hierarchical Spatiotemporal Feature Fusion Network (HSFF-Net). Specifically, Bi-directional Temporal-Spatial Pyramid (BiTSFP), first application bi-directional fusion architectures field, which adds flow shallow location...

Cyber-Physical Systems (CPS) are hybrid, component-based and open systems. Hybrid interface automata (HIA), which extend from automata, introduced to model CPS. HIA not input-enable, that is, only certain inputs can be accepted on a location of HIA. Thus, specify assumptions about the environment made by component. use an optimistic approach composition, two components compatible if there is in they work together. It important describe continuous behavior as well discrete system. We give...

This Innovate Practice Full Paper describes our experience with teaching cybersecurity topics using guided inquiry collaborative learning. The goal is to not only develop the students' in-depth technical knowledge, but also “soft skills” such as communication, attitude, team work, networking, problem-solving and critical thinking. paper reports developing Guided Inquiry Collaborative Learning materials on of firewall IPsec. Pre- post-surveys were conducted access effectiveness developed...

The images of motion states are time-varying, and when actually detecting their internal targets, the formed detection frames overlap, resulting in small confidence values for low accuracy results. To address this problem, authors propose a target image using improved YOLO algorithm. First, algorithm is deformable convolution; edge weights front back views within collated, segmented Second, structure by initial convolution used as frame structure, parallel cross-ratio value set, overlap...

Lane detection methods based on anchors have received increasing attention, but fixed-shape make it difficult to model complex lane line shapes. To solve this problem, we propose an Enhanced Curve Perception Network (ECPNet). Specifically, a Layer-by-layer Context Fusion (LCF) module fully utilize both high-level and low-level features in by establishing short hop connections across feature layers of diverse scales. Then, novel Structural Correction Prediction (SCP) module, which enhances...