检索结果-内蒙古大学图书馆

How Integration of a Brain-Machine Interface and Obstacle Detection System Can Improve Wheelchair Control via Movement imagery

引用

SENSORS 2024年第3期24卷 918页

作者： Kocejko, Tomasz Matuszkiewicz, Nikodem Durawa, Piotr Madajczak, Aleksander Kwiatkowski, Jakub Gdansk Univ Technol Fac Elect Telecommun & Informat Dept Biomed Engn Narutowicza 11-12 PL-80233 Gdansk Poland

This study presents a human-computer interaction combined with a brain-machine interface (BMI) and obstacle detection system for remote control of a wheeled robot through movement imagery, providing a potential solution for individuals facing challenges with conventional vehicle operation. The primary focus of this work is the classification of surface EEG signals related to mental activity when envisioning movement and deep relaxation states. Additionally, this work presents a system for obstacle detection based on image processing. The implemented system constitutes a complementary part of the interface. The main contributions of this work include the proposal of a modified 10-20-electrode setup suitable for motor imagery classification, the design of two convolutional neural network (CNNs) models employed to classify signals acquired from sixteen EEG channels, and the implementation of an obstacle detection system based on computer vision integrated with a brain-machine interface. The models developed in this study achieved an accuracy of 83% in classifying EEG signals. The resulting classification outcomes were subsequently utilized to control the movement of a mobile robot. Experimental trials conducted on a designated test track demonstrated real-time control of the robot. The findings indicate the feasibility of integration of the obstacle detection system for collision avoidance with the classification of motor imagery for the purpose of brain-machine interface control of vehicles. The elaborated solution could help paralyzed patients to safely control a wheelchair through EEG and effectively prevent unintended vehicle movements.

关键词： motor imagery brain-machine interface deep learning obstacle detection

来源：评论

学校读者我要写书评

暂无评论

AgriSen-COG, a Multicountry, Multitemporal Large-Scale Sentinel-2 Benchmark Dataset for Crop Mapping Using deep learning

引用

REMOTE SENSING 2023年第12期15卷 2980页

作者： Selea, Teodora West Univ Timisoara Fac Math & Informat Timisoara 300223 Romania

With the increasing volume of collected Earth observation (EO) data, artificial intelligence (AI) methods have become state-of-the-art in processing and analyzing them. However, there is still a lack of high-quality, large-scale EO datasets for training robust networks. This paper presents AgriSen-COG, a large-scale benchmark dataset for crop type mapping based on Sentinel-2 data. AgriSen-COG deals with the challenges of remote sensing (RS) datasets. First, it includes data from five different European countries (Austria, Belgium, Spain, Denmark, and the Netherlands), targeting the problem of domain adaptation. Second, it is multitemporal and multiyear (2019-2020), therefore enabling analysis based on the growth of crops in time and yearly variability. Third, AgriSen-COG includes an anomaly detection preprocessing step, which reduces the amount of mislabeled information. AgriSen-COG comprises 6,972,485 parcels, making it the most extensive available dataset for crop type mapping. It includes two types of data: pixel-level data and parcel aggregated information. By carrying this out, we target two computer vision (CV) problems: semantic segmentation and classification. To establish the validity of the proposed dataset, we conducted several experiments using state-of-the-art deep-learning models for temporal semantic segmentation with pixel-level data (U-Net and ConvStar networks) and time-series classification with parcel aggregated information (LSTM, Transformer, TempCNN networks). The most popular models (U-Net and LSTM) achieve the best performance in the Belgium region, with a weighted F1 score of 0.956 (U-Net) and 0.918 (LSTM).The proposed data are distributed as a cloud-optimized GeoTIFF (COG), together with a SpatioTemporal Asset Catalog (STAC), which makes AgriSen-COG a findable, accessible, interoperable, and reusable (FAIR) dataset.

关键词： benchmark dataset crop monitoring crop detection deep learning image segmentation multitemporal analysis crop classification agricultural application Sentinel-2 common agricultural policy (CAP)

来源：评论

学校读者我要写书评

暂无评论

Locating people in real-World for Assisting Crowd Behaviour Analysis Using SSD and deep SORT Algorithm 6

Locating people in Real-World for Assisting Crowd Behaviour ...

引用

6th International Conference on Wireless Communications, Signal processing and Networking (IEEE WiSPNET)

作者： Juyal, Piyush Sharma, Sachin Graph Era Deemed Univ Dept Comp Sci & Engn Dehra Dun Uttarakhand India

ISBN: (纸本)9781665440868

In security-sensitive areas, tracking individuals can be of high importance or may even be used to understand crowd activity, how people change their movement within the crowd or most common shops. Crowd activity trends can help to determine the most common locations of public places, which can help attract a wider audience in advertising placement. A growing concern has been public safety;it can be a harrowing experience to track these mischievous elements inside the crowd. This paper provides an approach that can map people with a geographical coordinate, which denotes an individual's realtime position in the real world through a camera feed. This data helps us to track the road, the direction of motion and where a person has turned on a map. These insights include descriptions of in-depth crowd movement that can assist with better surveillance and evaluating crowd activity.

关键词： Monitoring real-time systems image edge detection Visual analytics Data visualization deep learning

来源：评论

学校读者我要写书评

暂无评论

ICASSP 2021 deep NOISE SUPPRESSION CHALLENGE

ICASSP 2021 DEEP NOISE SUPPRESSION CHALLENGE

引用

IEEE International Conference on Acoustics, Speech and Signal processing (ICASSP)

作者： Reddy, Chandan K. A. Dubey, Harishchandra Gopal, Vishak Cutler, Ross Braun, Sebastian Gamper, Hannes Aichner, Robert Srinivasan, Sriram Microsoft Corp Redmond WA 98052 USA

ISBN: (纸本)9781728176055

The deep Noise Suppression (DNS) challenge is designed to foster innovation in the area of noise suppression to achieve superior perceptual speech quality. We recently organized a DNS challenge special session at INTERSPEECH 2020 where we open-sourced training and test datasets for researchers to train their noise suppression models. We also open-sourced a subjective evaluation framework and used the tool to evaluate and select the final winners. Many researchers from academia and industry made significant contributions to push the field forward. We also learned that as a research community, we still have a long way to go in achieving excellent speech quality in challenging noisy real-time conditions. In this challenge, we expanded both our training and test datasets. Clean speech in the training set has increased by 200% with the addition of singing voice, emotion data, and non-English languages. The test set has increased by 100% with the addition of singing, emotional, non-English (tonal and non-tonal) languages, and, personalized DNS test clips. There are two tracks with focus on (i) real-time denoising, and (ii) real-time personalized DNS. We present the challenge results at the end.

关键词： Speech Enhancement Perceptual Speech Quality P.808 deep Noise Suppressor DNS Machine learning

来源：评论

学校读者我要写书评

暂无评论

基于双视角CT重建的固体发动机点火试验药柱缺陷实时检测技术

引用

固体火箭技术 2023年第1期46卷 128-137页

作者：武晶陆明陈平潘晋孝中北大学信息与探测处理山西省重点实验室太原030051 中北大学理学院太原030051 中国航天科工集团有限公司六院601所呼和浩特010076

为了在固体火箭发动机点火试验环境下,研究药柱裂纹等缺陷对燃面退移的影响,需要对药柱进行实时缺陷检测。但由于传统CT重建算法和现有的基于深度学习的方法无法实时、有效地重建内部缺陷来进行动态监测,提出了一种利用双视角投影数据进... 详细信息

为了在固体火箭发动机点火试验环境下,研究药柱裂纹等缺陷对燃面退移的影响,需要对药柱进行实时缺陷检测。但由于传统CT重建算法和现有的基于深度学习的方法无法实时、有效地重建内部缺陷来进行动态监测,提出了一种利用双视角投影数据进行CT重建来检测缺陷的方法。首先,该方法利用深度学习模型进行双视角下发动机内结构重建,完成二维到三维的跨维度重建;然后,利用无监督异常检测算法对双视角下投影图像进行缺陷检测及定位,获取缺陷区域坐标;最后,利用视差原理在三维结构体中重建缺陷,从而可以实时得到发动机内部三维结构,达到动态监测目的。利用仿真数据与模拟工件数据进行实验,实验结果表明,该方法仅利用双视角下的投影图像即可有效完成三维重建以进行缺陷检测。相比于现有的基于深度学习的超稀疏重建算法,无需缺陷区域先验信息。

关键词：发动机药柱缺陷实时检测双视角CT重建深度学习视差原理

来源：评论

学校读者我要写书评

暂无评论

High-Speed Tracking with Mutual Assistance of Feature Filters and Detectors

引用

SENSORS 2023年第16期23卷 7082页

作者： Matsuo, Akira Yamakawa, Yuji Univ Tokyo Grad Sch Interdisciplinary Informat Studies Tokyo 1538505 Japan Univ Tokyo Interfac Initiat Informat Studies Tokyo 1538505 Japan Univ Tokyo Inst Ind Sci 4-6-1 KomabaMeguro ku Tokyo 1538505 Japan

Object detection and tracking in camera images is a fundamental technology for computer vision and is used in various applications. In particular, object tracking using high-speed cameras is expected to be applied to real-time control in robotics. Therefore, it is required to increase tracking speed and detection accuracy. Currently, however, it is difficult to achieve both of those things simultaneously. In this paper, we propose a tracking method that combines multiple methods: correlation filter-based object tracking, deep learning-based object detection, and motion detection with background subtraction. The algorithms work in parallel and assist each other's processing to improve the overall performance of the system. We named it the "Mutual Assist tracker of feature Filters and Detectors (MAFiD method)". This method aims to achieve both high-speed tracking of moving objects and high detection accuracy. Experiments were conducted to verify the detection performance and processing speed by tracking a transparent capsule moving at high speed. The results show that the tracking speed was 618 frames per second (FPS) and the accuracy was 86% for Intersection over Union (IoU). The detection latency was 3.48 ms. These experimental scores are higher than those of conventional methods, indicating that the MAFiD method achieved fast object tracking while maintaining high detection performance. This proposal will contribute to the improvement of object-tracking technology.

关键词： high-speed vision image processing machine learning object tracking

来源：评论

学校读者我要写书评

暂无评论

Efficient and accurate object detection for 3D point clouds in intelligent visual internet of things

引用

MULtimeDIA TOOLS AND APPLICATIONS 2021年第20期80卷 31297-31334页

作者： Li, Hui Wang, Junyin Xu, Lingwei Zhang, Shujun Tao, Ye Qingdao Univ Sci & Technol Sch Informat Sci & Technol Qingdao 266061 Peoples R China South Cent Univ Nationalities Hubei Key Lab Intelligent Wireless Commun Wuhan 430074 Peoples R China Lanzhou Jiaotong Univ Minist Educ Key Lab Optotechnol & Intelligent Control Lanzhou 730070 Peoples R China

Visual perception is a key technology in the Intelligent Visual Internet of Things. The research of object detection methods is of great significance for improving the safety and efficiency of unmanned driving technology and intelligent visual Internet of Things. 3D point clouds object detection of deep learning can not only use the deep network to automatically learn characteristics of the multi-layer abstract structure, improve calculation efficiency and detection accuracy of the model, but also have better performance in dealing with object occlusion, absence and data sparsity with obtained high-dimensional point clouds information. However, the current review of object detection methods for 3D point clouds based on deep learning is scarce. In order to provide a more comprehensive understanding and understanding of the security and efficiency development of driverless technology, this paper is divided into the monocular camera, RGB-D image and LiDAR point cloud, according to the main data of the network model, and further subdivides according to the different use methods of the model. Analyze the performance of various model detection methods. This article also summarizes current commonly used 3D point clouds datasets of object detection, organizes and describes detection metrics of commonly used 3D point clouds, and discusses research challenges and development trends. The real-time performance of 3D point cloud object detection under the intelligent vision Internet of Things needs to be improved.

关键词： Internet of things Object detection 3D point clouds Monocular camera RGB-D image LiDAR

来源：评论

学校读者我要写书评

暂无评论

UAV Detection using Web Application Approach based on SSD Pre-Trained Model 5

UAV Detection using Web Application Approach based on SSD Pr...

引用

5th IEEE International Conference on Aerospace Electronics and Remote Sensing Technology (ICARES)

作者： Wastupranata, Leonard Matheus Munir, Rinaldi Inst Teknol Bandung Sch Elect Engn & Informat Bandung Indonesia

ISBN: (纸本)9781665401791

UAV development is being intensively developed by various groups to help overcome various types of problems. Object Detection is important in helping UAVs to do drone chasing and other competition that need visual approach based on image processing and deep learning. Unfortunately, the computational capabilities of the onboard processing unit that attached to the UAV are less than optimal for object detection due to storage and memory size constraints. This paper aims to create the new approach to improve the precision and recall during UAV detection by using web application to do real time detection. To decide a pre-trained model, it is necessary to compare which SSD pre-trained model is suitable to be deployed in this web application. The results obtained are that using the web application approach is better than the onboard processing approach with a high level of precision and recall with an average precision value of 0.85 and an average recall value of 0.837.

关键词： UAV object detection SSD web application deep learning

来源：评论

学校读者我要写书评

暂无评论

FCN-Based Carrier Signal Detection in Broadband Power Spectrum

引用

IEEE ACCESS 2020年 8卷 113042-113051页

作者： Huang, Hao Li, Jian-Qing Wang, Mao Wang, Hong Univ Elect Sci & Technol China Sch Elect Sci & Engn Chengdu 610054 Peoples R China

Carrier signal detection has been a problem for a long time, which is the first step for blind signal processing. In this paper, we propose a new method for carrier signal detection in the broadband power spectrum based on the fully convolutional network (FCN). FCN is a deep learning method and used in semantic image segmentation tasks. By regarding the broadband power spectrum sequence as a one-dimensional (1D) image and each subcarrier on the broadband as the target object, we can transform the carrier signal detection problem on the broadband into a semantic segmentation problem on a 1D image without prior knowledge. We design a 1D deep convolutional neural network (CNN) based on FCN to classify each point on broadband power spectrum array into two classes: subcarrier or noise, and then we can easily locate the subcarrier signals' position on the broadband power spectrum. We train the deep CNN on a simulation dataset and validate it on a real satellite broadband power spectrum dataset. The experimental results show that our method can effectively detect the subcarrier signal in the broadband power spectrum and achieve higher accuracy than the slope tracing method.

关键词： Carrier signal detection broadband power spectrum deep learning fully convolutional networks

来源：评论

学校读者我要写书评

暂无评论

FACE MASK DETECTION USING deep learning

FACE MASK DETECTION USING DEEP LEARNING

引用

11th International Conference of Computer Communication and Informatics (ICCCI)

作者： Kodali, Ravi Kishore Dhanekula, Rekha Natl Inst Technol Warangal Dept Elect & Commun Engn Warangal Andhra Pradesh India

ISBN: (纸本)9781728158754

In order to prevent the spread of CORONA virus, everyone must wear a mask during the pandemic. In these tough times of COVID-19 it is necessary to build a model that detects people with and without mask in real-time as it works as a simple precautionary measure to prevent the spread of virus. If deployed correctly, this machine learning technique helps in simplifying the work of frontline warriors and saving their lives. A basic Convolutional Neural Network (CNN) model is built using TensorFlow, Keras, Scikit-learn and OpenCV to make the algorithm as accurate as possible. Javascript API helps in accessing webcam for real-time face mask detection. Since Google Colab runs on web browser it can't access local hardware like a camera without APIs. The proposed work contains three stages: (i) pre-processing, (ii) Training a CNN and (iii) real-time classification. The first part is the Pre-processing section, which can be divided into "Grayscale Conversion" of RGB image, "image resizing and normalization" to avoid false predictions. Then the proposed CNN, classifies faces with and without masks as the output layer of proposed CNN architecture contains two neurons with Softmax activation to classify the same. Categorical cross-entropy is employed as loss function. The proposed model has Validation accuracy of 96%. If anyone in the video stream is not wearing a protective mask a Red coloured rectangle is drawn around the face with a dialog entitled as NO MASK and a Green coloured rectangle is drawn around the face of a person wearing MASK

关键词： Convolutional Neural Network (CNN) Corona Virus Face mask detection

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：