With the wide application of unmanned aerial vehicle (UAV), the problem of object detection in UAV aerial image has attracted more and more attention. In order to detect objects in UAV aerial images more accurately, w...
With the wide application of unmanned aerial vehicle (UAV), the problem of object detection in UAV aerial image has attracted more and more attention. In order to detect objects in UAV aerial images more accurately, we made a series of improvements on the basis of YOLOv7-tiny to make it more suitable for detecting objects in UA V aerial images. Specifically, we added Global Attention Mechanism (GAM) to YOLOv7-tiny's neck network to better extract the features of the objects. We also introduced the Bidirectional Feature Pyramid Network (BiFPN) into the Neck network to enhance feature fusion capability. In order to detect small objects more accurately, we added a small object detection head in the YOLOv7-tiny Head section. Contextual Transformer module was introduced to make the network pay more attention to context information. To make the predicted box match the ground truth box more closely, we applied the SIoU loss function to YOLOv7-tiny. The experimental results showed that, mAP@0.5 is 38.1 % and mAP@0.5:0.95 is 21.3% by improved YOLOv7-tiny algorithm on VisDrone dataset. Compared with YOLOv7-tiny, mAP@0.5 is improved by 2.8% and mAP@0.5:0.95 is improved by 3%, which improved the accuracy of object detection in UAV aerial images.
Thelandscape of mobile communication networks has been evolving so *** the last decades,3G/4G achieved unprecedented success thanks to its advancement in IP for all,and high-speed,mobile and ubiquitous Internet *** wi...
Thelandscape of mobile communication networks has been evolving so *** the last decades,3G/4G achieved unprecedented success thanks to its advancement in IP for all,and high-speed,mobile and ubiquitous Internet *** with 3G/4G,5G now makes significant advances in terms of connectivity,data rate and latency,which shifts from the conventional people-to-people to the modern people-to-things and things-to-things communication *** the next standardization focus,6G will further build a ubiquitous network that realizes the integrated communication of air,space,earth and sea,boost the data rate 100 X faster than that of 5G,enable machine intelligence through AI,and many ***,6G will have much broader appli-cation prospects such as virtual or augmented reality,Internet of Things,Industrial Internet,unmanned driving,and smart factories.
Cesium lead iodide (CsPbI3) perovskite has gained considerable attentions due to its unique photoelectric properties. However, numerous defects are inevitably existed on the surface and grain boundaries (GBs) of CsPbI...
详细信息
Due to hardware constraints, no satellite sensor can supply images with all high spatial, high temporal and high spectral resolutions nowadays. Spatialtemporal-spectral fusion (STSF) is supposed as an economical and f...
详细信息
Non-intrusive Load Monitoring (NILM) proves effective in disaggregating energy consumption data for individual household appliances. Traditional NILM techniques rely on one-dimensional (1D) time series signal (TSS) an...
详细信息
ISBN:
(数字)9798350318098
ISBN:
(纸本)9798350318104
Non-intrusive Load Monitoring (NILM) proves effective in disaggregating energy consumption data for individual household appliances. Traditional NILM techniques rely on one-dimensional (1D) time series signal (TSS) analysis to predict the energy usage of distinct appliances. Nonetheless, these approaches have limitations in handling data with non-linear behavior and complex patterns. To address this challenge, prior studies have proposed converting 1D TSS into two-dimensional (2D) images using techniques like Gramian Angular Field (GAF), Markov Transition Field (MTF), and Recurrence Plot (RP). Recent research underscores the effectiveness of the Gramian Angular Summation Field (GASF) and Gramian Angular Difference Field (GADF), both variants of GAF, in the NILM domain. This study assesses the performance of three distinct 1D to 2D conversion approaches for energy disaggregation: GADF, GASF, and an enhanced hybrid method integrating GADF and GASF. Performance evaluation utilizes matrices, including accuracy, precision, recall, and the F1 score. The experiments are conducted on a publicly accessible real-world energy dataset encompassing different appliances. The results demonstrate that the combined GADF and GASF method introduced in this study outperforms their individual counterparts with better performance and faster convergence, particularly in the context of energy disaggregation. The findings offer valuable insights for advancing the development of more accurate and efficient NILM techniques.
By 2050, experts predict that 75% of people on Earth will live in cities. As such, cities need to be prepared to fulfill the needs of their citizens and provide the best possible services. The idea of a “smart city,...
By 2050, experts predict that 75% of people on Earth will live in cities. As such, cities need to be prepared to fulfill the needs of their citizens and provide the best possible services. The idea of a “smart city,” which is an improved system that effectively manages its resources and services via monitoring and communication technology, is often central to the notion of an advanced city. Thus, turning cities into smart entities is a critical step toward global urban sustainability. Beyond creating smart buildings in cities, the main goal is to improve people's quality of life by using newly developed digital technologies and sensing capabilities. The key to smarter communities is their capacity to invest in these technologies, adapt, and socialize in a transparent and inclusive manner while taking into account the needs and values of the local and regional communities. Despite these developments, privacy issues and interruptions in cyber security remain serious weaknesses that need addressed.
The paper discusses the topic of determining the current-voltage characteristics of two-terminal passive components. A virtual measurement instrument created in the LabVIEW environment using the NI USB-6211 data acqui...
The paper discusses the topic of determining the current-voltage characteristics of two-terminal passive components. A virtual measurement instrument created in the LabVIEW environment using the NI USB-6211 data acquisition device is presented, as well as the results of tests performed for example elements . The described approach can also be applied to other measurement situations, such as for studying the self-heating effect of RTD.
This paper presents a feasibility study of a pH sensor utilizing the microwave transmission technique within the CST Microwave Studio software. The resonator is designed based on the modified split ring resonator shap...
This paper presents a feasibility study of a pH sensor utilizing the microwave transmission technique within the CST Microwave Studio software. The resonator is designed based on the modified split ring resonator shape (SRR) by incorporating six segments in a circular shape. The proposed resonator shape resembles a honeycomb structure with a capacitive gap to enhance the electric field (E-field) distribution in its area. The operating frequency is set to the Industrial, Scientific, and Medical (ISM) band of 5.82 GHz. According to the simulation, the E-field demonstrates its maximum strength at the focused area, measuring 29,863 V/m. Consequently, this position is deemed suitable for selecting the sensing area for liquid-mixture detection. To investigate the performance of the proposed sensor, the industry model of a Teflon tube is imported and placed atop the focused honeycomb resonator. The performance of the proposed sensor is evaluated by modeling various liquid samples, adjusting electrical properties such as relative permittivity, εr, and loss factor. The simulated results indicate that the proposed sensor provides a satisfactory transmission coefficient, S21, response for classifying relative permittivity, εr, changes within the range of 1 to 81. However, for future development, the proposed sensor will undergo fabrication and further investigation to assess its performance in practical applications.
Motion direction detection is an important part of the visual system. In the past few decades, research on the directional selectivity of motion detection is of great progress. However, they are all concentrated on th...
详细信息
暂无评论