Intelligence is one of the most important aspects in development our future communities. Ranging from smart home to building city, all these infrastructures must be supported by intelligent power supply. Smart grid proposed solve challenges electricity In grid, order realize optimal scheduling, an SM installed at each collect near-real-time consumption data, which can used utilities offer better services. However, data may disclose a user's private information. An adversary track application...

In the field of object detection, recently, tremendous success is achieved, but still it a very challenging task to detect and identify objects accurately with fast speed. Human beings can recognize multiple in images or videos ease regardless object’s appearance, for computers distinguish between things. this paper, modified YOLOv1 based neural network proposed detection. The new model has been improved following ways. Firstly, modification made loss function network. replaces margin style...

Laser-targeted weeding methods further enhance the sustainable development of green agriculture, with one key technology being improvement weed localization accuracy. Here, we propose an improved YOLOv8 instance segmentation based on bidirectional feature fusion and deformable convolution (BFFDC-YOLOv8-seg) to address challenges insufficient accuracy in complex environments resource-limited laser devices. Initially, by training extensive datasets plant images, most appropriate model scale...

It is of great importance for procedure retrieval to find an effective classification method Chinese legal documents with deep semantic understanding, as the electronic law have massive volume and complex structure. In this paper, a learning document using Graph LSTM (Long Short-Term Memory) combined domain knowledge extraction proposed. First, judicial model constructed based on ontologies that include top-level ontology domain-specific ontology. Second, are divided into different blocks...

Ontology reuse offers great benefits by measuring and comparing ontologies. However, the state of art approaches for ontologies neglects problems both polymorphism ontology representation addition implicit semantic knowledge. One way to tackle these is devise a mechanism measurement that stable, basic criteria automatic measurement. In this paper, we present graph derivation based approach (GDR) stable measurement, which captures structural semantics addresses those cause unstable This paper...

Internet of Energy is considered as a promising approach to solve the problems energy crisis and carbon emission. It needs collect user's real-time data for optimizing utilization. However, such may disclose privacy information. Previous works usually adopt specific obfuscation value mask counteract deviation through aggregation; these can preserve effectively, but most them consider less about data-utility (precision). In this paper, we propose utility-privacy tradeoff scheme based on...

This work describes the electromechanical behaviour of iron- and silver-sheathed ex situ Sr0.6Ka0.4Fe2As2 (Sr-122) tape superconductors. Free-standing Sr-122 tapes were stressed by axial tension at 4.2 K under external fields 5–6 T, critical currents have been simultaneously measured. An attempt to reinforce Sr-122/Ag mechanically was done simple soldering two stainless steel strips 40 μm both sides tape. The obtained results are compared with property MgB2/Fe. mechanical limits Sr-122/Fe...

Laser-targeted weeding methods further enhance the sustainable development of green agriculture, with one key technology being improvement weed localization accuracy. Here, we propose an improved YOLOv8 instance segmentation based on bidirectional feature fusion and deformable convolution (BFFDC-YOLOv8-seg) to address challenges insufficient accuracy in complex environments resource-limited laser devices. Initially, by training extensive datasets plant images, most appropriate model scale...

Despite the recent dramatic advances in object detection, detecting a small general and remote sensing images is still challenging problem. One main reason for this appearance of objects images. Specifically low resolution noisy representation makes it hard to detect objects. We tickle down problem by proposing novel detector based on Generative adversarial network (GAN), which we called FPN-GAN short. The proposed method composed GAN, Resnet-50 as backbone, Feature Pyramid Network...

In this paper, we propose a set of semantic cohesion metrics for ontology measurement, which can be used to assess the quality evolving ontologies in context dynamic and changing Web. We argue that these are stable measure semantics rather than structures. Measuring with inconsistencies caused by evolutions, is mainly considered paper. The proposed theoretically empirically validated.

Abstract With the drastic development of semantic‐driven applications, assessing quality ontologies has received more attention. Measuring and can help ontology engineers to control project management reduce risk failures. However, most existing metrics for measuring are defined based on structure, neglect stability measurement. In this paper, we concentrate stable measurement by using semantically derived metrics. We propose four cohesion metrics, which fully consider implicitly expressed...

Single Shot Multibox Detector (SSD), which is considered one of the top prominent algorithms for object detection in terms speed and accuracy. However, feature pyramid, conventional SSD uses every layer individualistically ignores background information objects by taking only fine-grained details into account, drops its accuracy certain cases especially heavy occlusion, overlapping small objects. It a known fact that if number maps increased, efficiency deep network to be but just simply...