检索结果-内蒙古大学图书馆

LCDNet: Lightweight Change Detection Network With Dual-Attention Guidance and Multiscale Feature Fusion for Remote-Sensing images

引用

IEEE GEOSCIENCE AND REMOTE SENSING LETTERS 2024年 21卷 1页

作者： Li, Junwei Li, Shijie Wang, Feng Henan Univ Sch Artificial Intelligence Zhengzhou 450046 Peoples R China Weinan Normal Univ Sch Phys & Elect Engn Weinan 714099 Peoples R China

deep-learning (DL)-based change detection (CD) techniques have recently become increasingly complex to produce more accurate detection results. However, the increase in complexity leads to reduced efficiency and limits the application of DL-based CD techniques in domains that require real-time performance. To this end, a lightweight CD network (LCDNet) is proposed to accurately recognize changes in remote-sensing (RS) image pairs while maintaining high efficiency. First, a focus module is utilized at the beginning of the encoding layer for the downsampling operation, which reduces the computation of the model and the loss of information. Then, a depthwise (DW) convolution-based efficient extraction block (EEB) is designed by stacking different sizes of convolution kernels for the effective extraction of change features under different receptive fields. Next, a dual-attention guidance module (DAGM) is designed to guide the encoder in processing and selectively aggregating information related to changes. Lastly, a multiscale feature fusion module (MFFM) with low parameters is proposed that combines feature maps of different scales to exploit their complementary information. Compared with other state-of-the-art (SOTA) methods, the proposed LCDNet only requires approximately 0.83 M Params, 2.03 G FLOPs, and 3.03 ms inference time (It) to remarkably surpass them in terms of accuracy. Moreover, compared with other dual-attention and multiscale fusion modules, the proposed DAGM and MFFM are more effective and efficient. The source code will be made available at https://***/sjl2023/LCDNet.

关键词： Convolution Feature extraction Training Kernel Measurement Remote sensing image coding Change detection (CD) lightweight multiscale features remote-sensing (RS)

来源：评论

学校读者我要写书评

暂无评论

ISPA 2021 - 12th International Symposium on image and Signal processing and Analysis

ISPA 2021 - 12th International Symposium on Image and Signal...

引用

12th International Symposium on image and Signal processing and Analysis, ISPA 2021

ISBN: (纸本)9781665426398

The proceedings contain 47 papers. The topics discussed include: a real time subband implementation of an active noise control system for snoring reduction;ultrasound anomaly detection based on variational autoencoders;a sparse TFD reconstruction approach using the s-method and local entropies information;illuminant estimation error detection for outdoor scenes using transformers;rapid defect detection by merging ultrasound B-scans from different scanning angles;mushroom image classification with CNNs: a case-study of different learning strategies;epsilon greedy strategy for hyper parameters tuning of a neural network equalizer;automated ultrasonic testing of materials based on C-scan flaw classification;and robust deep simple online real-time tracking.

关键词：

来源：评论

学校读者我要写书评

暂无评论

deep learning-Based Human Activity Recognition Algorithms: A Comparative Study

Deep Learning-Based Human Activity Recognition Algorithms: A...

引用

image processing and Computer Applications (ICIPCA), IEEE International Conference on

作者： Zhenghui Lu Southern Utah University Guangzhou Guangdong China

Human activity recognition is essential for developing personalized healthcare, fitness tracking, and fall detection systems. This study focuses on developing an accurate deep learning-based algorithm for recognizing human activities using accelerometer and gyroscope data. The proposed model is based on a 1D-CNN-BiLSTM architecture and is trained on a public dataset. In this study, six different human actions-walking, walking upstairs, walking downstairs, sitting, standing, and laying-are identified to examine how well the proposed model performs compared to a baseline model and to assess how well the model recognizes these activities. According to the results, the suggested model performs better than the baseline model at identifying human activity and is more accurate overall. All six activities can be recognized by the suggested model with high accuracy, with walking, walking downstairs, and walking upwards having the best performance. The model's success is attributed to its ability to capture both spatial and temporal dependencies in the data, allowing it to effectively recognize human activities based on accelerometer and gyroscope data. The study's findings have significant implications for developing activity recognition systems, especially for real-time health monitoring, fall detection, and fitness tracking applications. The proposed model's high accuracy can improve the performance of such systems, leading to better outcomes for patients and individuals. The study's limitations are also discussed, including the relatively small size of the dataset and the lack of variability in the activities performed. This study provides evidence of the effectiveness of a deep learning-based algorithm for recognizing human activities using accelerometer and gyroscope data. It demonstrates the potential of such models for developing accurate and reliable activity recognition systems with important applications in healthcare and fitness tracking.

关键词：

来源：评论

学校读者我要写书评

暂无评论

EyeRis: Visual image Recognition using Machine learning for the Visually-Impaired

EyeRis: Visual Image Recognition using Machine Learning for ...

引用

International Conference on Electronics, Information and Communications (ICEIC)

作者： Alliah May Eugenio Marian Jowie Patulot Lykha Jane Seguiro Angelica Nicole Tuazon Shekinah Lor Huyo-a Mideth Abisado Gabriel Avelino Sampedro Department of Computer Science Technological University of the Philippines Manila Philippines Philippine Coding Camp Research and Development Center Manila Philippines College of Computing and Information Technologies National University Manila Philippines Faculty of Information and Communication Studies University of the Philippines Open University Laguna Philippines

Visually impaired people struggle daily and have difficulty recognizing and distinguishing objects around them. Thus, they mainly depend on supervision from other people to assist them. Since smartphones have become a necessity in this modern world, the researchers formulated a solution to help the visually impaired through a machine learning-based mobile application for object recognition. Nowadays, software applications can provide accurate findings in picture classification and processing processes thanks to machine learning techniques and algorithms. In this study, the researchers use a convolutional neural network (CNN)-based system on TensorFlow Lite to create a mobile version of a visual information system employing a machine learning strategy and deep learning framework. The main objectives of the smartphone application, EyeRis, is to recognize and categorize items in real-time and to separate photographs from the user-selected scenarios. Results were analyzed and contrasted based on the app’s obtained recognition accuracy data. It demonstrated the utility of CNN as a model for image recognition algorithms.

关键词： Visualization image recognition Machine learning algorithms Software algorithms Neural networks Software real-time systems

来源：评论

学校读者我要写书评

暂无评论

Ensemble learning based approach for traffic incident detection and multi-category classification

引用

ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE 2024年 132卷

作者： Xu, Miao Liu, Hongfei Yang, Hongbo Jiangsu Univ Sch Automot & Traff Engn Zhenjiang 212013 Peoples R China Jilin Univ Coll Transportat Changchun 130022 Peoples R China

Traffic incident is one of the important causes of road congestion. Traffic incident detection plays a crucial role in the safety application of intelligent transportation systems, which provides timely information for traffic management departments and reducing losses. Despite many researches on incident detection approach, the identification of different incident categories is not enough. In addition, traffic incident detection is still a challenging task due to the problem of data imbalance and feature selection. In this study, we propose a two-stage traffic incident detection framework based on ensemble learning. In the first stage, a binary classification algorithm based on XGBoost (eXtreme Gradient Boosting) is established to detect whether there is a traffic incident, and 24 feature variables are determined by model feature selection. In the second stage, three resampling algorithms are utilized to reconstruct and balance the dataset. Through comparative analysis, SMOTE (Synthetic Minority Oversampling Technique)-XGBoost is the best method for incident multi-category classification with precision of 87.27%, 78.52% and 92.54%, respectively. Moreover, the baseline comparison experiments are conducted to evaluate our model performance with real-word datasets. The proposed model achieves the highest average accuracy of 93.45% in the first stage and the macro-precision of 86.11% in the second stage. The results indicate that the proposed method outperforms baselines and the two-stage framework can accurately realize the incident detection and multi-category classification.

关键词： Traffic incident detection Multi classification Ensemble learning Boosting algorithm Data analysis

来源：评论

学校读者我要写书评

暂无评论

Analysis of Stable Diffusion-derived fake weeds performance for training Convolutional Neural Networks

引用

COMPUTERS AND ELECTRONICS IN AGRICULTURE 2023年 214卷

作者： Moreno, Hugo Gomez, Adria Altares-Lopez, Sergio Ribeiro, Angela Andujar, Dionisio CSIC Ctr Automat & Robot Ctra Campo Real Km 0-200 La Poveda Arganda Del Rey 28500 Madrid Spain

Weeds challenge crops by competing for resources and spreading diseases, impacting crop yield and quality. Effective weed detection can enhance herbicide application, thus reducing environmental and health risks. A major challenge in Site-Specific Weed Management (SSWM) is developing a reliable weed identification system, especially given the diversity and similarity between certain weeds and crops during early growth stages. image-based deep learning (DL) methods have become vital for weed classification. However, accurate weed classification and detection using DL techniques face the bottleneck of requiring large labeled data. Furthermore, labeling this specific extensive data is a time-consuming and tedious task apart from necessitating weed science experts. This research's central focus is to present a novel approach to weed detection using convolutional neural network (CNN) classifiers, specifically Yolov8l and RetinaNet, augmented with Stable Diffusion data i.e., artificial weed images. Stable Diffusion enhanced the training data, increasing the classifiers' adaptability. The study targeted specific weeds (Solanum nigrum L.;Portulaca oleracea L.;Setaria Verticillata L.) found in tomato crops, using a limited number of real images (30 samples) to produce artificial training images for the CNNs. All validation and test sets are comprised of real weed images. Results showed high performance when using only artificial images in terms of Mean Average Precision (mAP). In isolated conditions (0.91 mAP), i.e., only one weed species per image, an average performance gain of about 3% in all tests is obtained. When adding the artificial images to the real ones (mixed dataset), a mAP of 0.99 is obtained. In contrast, results using only artificial images obtained 0.81 mAP when detecting more than a single weed species. However, when implementing the trained CNNs with a mixed dataset, a 6% -9% performance gain was achieved in all cases. A mAP of up to 0.93 was achieved in t

关键词： Stable diffusion Artificial dataset Weeds identification deep learning Neural networks Object detection image Robustness-data augmentation

来源：评论

学校读者我要写书评

暂无评论

SLIMBRAIN: Augmented reality real-time acquisition and processing system for hyperspectral classification mapping with depth information for in-vivo surgical procedures

引用

JOURNAL OF SYSTEMS ARCHITECTURE 2023年 140卷

作者： Sancho, Jaime Villa, Manuel Chavarrias, Miguel Juarez, Eduardo Lagares, Alfonso Sanz, Cesar Univ Politecn Madrid Res Ctr Software Technol & Multimedia Syst Madrid Spain Univ Complutense Madrid Hosp Univ 12 Octubre Dept Neurosurg Madrid Spain

Over the last two decades, augmented reality (AR) has led to the rapid development of new interfaces in various fields of social and technological application domains. One such domain is medicine, and to a higher extent surgery, where these visualization techniques help to improve the effectiveness of preoperative and intraoperative procedures. Following this trend, this paper presents SLIMBRAIN, a real-time acquisition and processing AR system suitable to classify and display brain tumor tissue from hyperspectral (HS) information. This system captures and processes HS images at 14 frames per second (FPS) during the course of a tumor resection operation to detect and delimit cancer tissue at the same time the neurosurgeon operates. The result is represented in an AR visualization where the classification results are overlapped with the RGB point cloud captured by a LiDAR camera. This representation allows natural navigation of the scene at the same time it is captured and processed, improving the visualization and hence effectiveness of the HS technology to delimit tumors. The whole system has been verified in real brain tumor resection operations.

关键词： Brain tumor AR GPU Machine learning Medicine Depth image Lidar Classification IA Neurosurgery Biomedical engineering

来源：评论

学校读者我要写书评

暂无评论

Navigation and Task Planning of a Mobile Robot under ROS Environment: A Case Study Using AutoRace Challenge 60

Navigation and Task Planning of a Mobile Robot under ROS Env...

引用

60th Annual Conference of the Society-of-Instrument-and-Control-Engineers-of-Japan (SICE)

作者： Shen, Ding-Ruei Chin, Hong-Liang Tu, Chih-Hung Chih, Jhao-Sian Venglar, Vojtech Chen, Kuo-Shen Krejsa, Jiri Vechet, Stanislav Natl Cheng Kung Univ Dept Aeronaut & Astronaut Tainan Taiwan Natl Cheng Kung Univ Dept Mech Engn Tainan Taiwan Brno Univ Technol Fac Mech Engn Brno Czech Republic

ISBN: (纸本)9784907764739

With the development in artificial intelligence, autonomous vehicle or mobile robot has gradually shown their potentials in various applications. Therefore, various contest platforms, such as Dockietown and AutoRace, have been developed for learning related techniques. This paper summarizes and demonstrates our efforts on developing various navigation and task planning missions for the AutoRace challenge, in which mobile robots must recognize a designated track, to reach the destination with minimum time and numbers of faults. In this work, a ROS-based mobile robot: Turtlebot Burger, is hired as the development platform and contest vehicle. In cooperation with cameras and a 2D lidar, various image processing techniques and deep learning algorithms are developed and coded for accomplishing the missions. The developed system could be split into three main parts, which are responsible for system management, image processing, and decision making for missions, respectively. The details of their relationships and the working flow will be further discussed in this paper. Also, a deep learning object detection model, YoloV4, is introduced to improve the detection of traffic signs. Finally, through the spent efforts, the entire mission can be smoothly accomplished within 2 to 3 minutes under normal situations. In the end, we will also discuss some problems caused by the limitations of hardware and system capability. Currently, efforts for improving the success rate in various situations are investigated to further reduce the time and increase stability. These results would be further studied for application in real-life technologies.

关键词： Mobile robots Navigation deep learning ROS Artificial intelligence

来源：评论

学校读者我要写书评

暂无评论

Optimized YOLOv8 (YOLO-Prime), M-CNN, and Dynamic Linear Regression for real-time Aerial Vehicle Detection in Embedded Systems

Optimized YOLOv8 (YOLO-Prime), M-CNN, and Dynamic Linear Reg...

引用

Human-Computer Interaction, Optimization and Robotic Applications (HORA), International Congress on

作者： Halil Hüseyin Çalışkan Talha Koruk Department of Computer Engineering Bursa Technical University Bursa Turkey

ISBN: (数字)9798331510886

ISBN: (纸本)9798331510893

This paper introduces a modified architecture of YOLOv8, called YOLO-Prime, to detect aerial vehicles. By employing YOLO-Prime to embedded systems such as Nvidia Jetson Xavier Nx and Rockchip 3588, we reduced the inference time by 130% compared to the YOLOv8m model and by 158% compared to the YOLOv8l model. In addition to YOLO-Prime, we present a Morphological CNN (M-CNN) to classify regions detected by YOLO-Prime. With the integration of M-CNN, we achieved an increase in detection accuracy by 20% in the detection of aerial vehicles. Furthermore, we introduce a dynamic linear regression model to predict the future coordinates of aerial vehicles. As a result of the integration of dynamic linear regression models to embedded systems, GPU utilization of deep learning models decreased by approximately 66%. In summary, the combination of YOLO-Prime, M-CNN, and the dynamic linear regression model contributes to detecting aerial vehicles in embedded systems where energy consumption, inference time, and accuracy are crucial for realtime applications.

关键词： YOLO Embedded systems Accuracy Robot kinematics image processing Linear regression Predictive models real-time systems Vehicle dynamics Optimization

来源：评论

学校读者我要写书评

暂无评论

Human Exclusive Alerts in CCTV Surveillance Using Computer Vision

Human Exclusive Alerts in CCTV Surveillance Using Computer V...

引用

2025 International Conference on Emerging Technologies for Intelligent Systems, ETIS 2025

作者： Varghese, Ann Mary Biju, Jeevan Manasrayil, Noble Jemini Shaji, Lisa Lakshmi, G. Thomas, Renjith Sjcet Palai Department of Ai & Ds Kottayam India Stcet Chengannur Department of Ai & Ml Kottayam India

ISBN: (纸本)9798331507541

This paper outlines a proposed approach to enhance modern surveillance systems by addressing the challenges posed by the excessive data generated by extensive CCTV camera networks. The proliferation of surveillance footage often results in inefficiencies, primarily due to false alarms triggered by non-human movements such as animals or environmental factors. The proposed system aims to tackle this issue by integrating advanced computer vision technologies with state-of-the-art deep learning algorithms, designed to accurately identify human presence and thereby significantly reduce irrelevant alerts Our intended solution involves the adoption of the latest YOLO (You Only Look Once) object detection model, YOLOv8, which is renowned for its efficiency and accuracy in real-time object recognition. We plan to combine this with OpenCV, a robust tool for image processing and computer vision tasks, to enhance our system's capability to process and analyze video data effectively. Furthermore, we aim to utilize the COCO dataset, a comprehensive object detection, segmentation, and captioning dataset, to train our deep learning algorithms. This training will equip our system to differentiate effectively between human and non-human elements by famil-iarizing it with a wide range of human poses and activities. The methodology we propose is currently undergoing rigorous validation and extensive real-world testing. We anticipate that these efforts will demonstrate the effectiveness of our approach in minimizing false alarms and maximizing the accuracy of human detection across various surveillance environments. By refining alert precision, our approach aims to enhance security measures, allowing surveillance operators to concentrate more on genuine human-related incidents. The research not only seeks to improve the operational efficiency of surveillance systems but also aims to contribute to the broader field of security technology by demonstrating potential applications of cutting

关键词： deep learning

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：