检索结果-内蒙古大学图书馆

Predicting the effective mechanical property of heterogeneous materials by image based modeling and deep learning

COMPUTER methods IN APPLIED MECHANICS AND ENGINEERING 2019年 347卷 735-753页

作者： Li, Xiang Liu, Zhanli Cui, Shaoqing Luo, Chengcheng Li, Chenfeng Zhuang, Zhuo Tsinghua Univ Appl Mech Lab Dept Engn Mech Sch Aerosp Beijing 100084 Peoples R China Swansea Univ Zienkiewicz Ctr Computat Engn Swansea W Glam Wales

In contrast to the composition uniformity of homogeneous materials, heterogeneous materials are normally composed of two or more distinctive constituents. It is usually recognized that the effective material property of a heterogeneous material is related to the mechanical property and the distribution pattern of each forming constituent. However, to establish an explicit relationship between the macroscale mechanical property and the microstructure appears to be complicated. On the other hand, machine learning methods are broadly employed to excavate inherent rules and correlations based on a significant amount of data samples. Specifically, deep neural networks are established to deal with situations where input-output mappings are extensively complex. In this paper, a method is proposed to establish the implicit mapping between the effective mechanical property and the mesoscale structure of heterogeneous materials. Shale is employed in this paper as an example to illustrate the method. At the mesoscale, a shale sample is a complex heterogeneous composite that consists of multiple mineral constituents. The mechanical properties of each mineral constituent vary significantly, and mineral constituents are distributed in an utterly random manner within shale samples. Large quantities of shale samples are generated based on mesoscale scanning electron microscopy images using a stochastic reconstruction algorithm. image processing techniques are employed to transform the shale sample images to finite element models. Finite element analysis is utilized to evaluate the effective mechanical properties of the shale samples. A convolutional neural network is trained based on the images of stochastic shale samples and their effective moduli. The trained network is validated to be able to predict the effective moduli of real shale samples accurately and efficiently. Not limited to shale, the proposed method can be further extended to predict effective mechanical properties o

关键词： Heterogeneous materials Shale Deep learning stochastic reconstruction

来源：评论

学校读者我要写书评

暂无评论

Convolutional neural Network Based Sound Recognition methods for Detecting Presence of Amateur Drones in Unauthorized Zones 2nd

Convolutional Neural Network Based Sound Recognition Methods...

引用

2nd International Conference on Machine Learning, image processing, Network Security and Data Sciences, MIND 2020

作者： Ganapathi, Ungati Sabarimalai Manikandan, M. Real-Time Embedded Signal Processing Lab School of Electrical Sciences Indian Institute of Technology Bhubaneswar Jatani Khordha752050 India

ISBN: (纸本)9789811563171

Unmanned aerial vehicles (UAVs), or drones, have become an integral part of diverse civil and commercial applications. But illegal operations of UAVs pose serious risks to public safety, privacy and national security. Thus, the detection of these vehicles has become vital to identify and track amateur drones in unauthorized zones or restricted areas. This paper presents three sound event recognition (SER) methods developed based on the Mel-frequency cepstrum coefficients (MFCCs) and spectrogram features combined with three machine learning classifiers such as multi-class SVM, one-dimensional convolutional neural networks (1DCNN) and two-dimensional CNN (2DCNN) for recognizing drone sounds. The SER methods are evaluated using a wide variety of sounds such as music, speech, wind, rain, and vehicle. Results showed that the MFCC-SVM based and MFCC-1DCNN based SER methods had average recall rate (RR) = 92.53%, precision rate (PR) = 93.21% and F1-score = 92.84% and RR = 94.28%, PR = 96.57% and F1-score = 95.02% for a segment length of 1 s, respectively. The spectrogram-2DCNN based method had average RR = 73.27%, PR = 74.55% and F1-score = 73.37% for audio segment length of 500 ms. Preliminary results demonstrate that the MFCC-1DCNN based SER method achieves better recognition rates as compared to that of the MFCC-SVM based and spectrogram-2DCNN based methods in recognizing drone sounds. © 2020, Springer Nature Singapore Pte Ltd.

关键词： Drones

来源：评论

学校读者我要写书评

暂无评论

基于偏振图像的低照度场景多目标检测算法

引用

红外技术 2022年第5期44卷 483-491页

作者：寻华生张晶晶刘晓李腾年福东张馨安徽大学计算智能与信号处理教育部重点实验室安徽合肥230601 偏振光成像探测技术安徽省重点实验室安徽合肥230031 中国科学院通用光学定标与表征技术重点实验室安徽合肥230031 合肥学院先进制造工程学院安徽合肥230601

偏振光反射信息可直接反演目标本征特性,且在传输过程中具备较强的抗干扰特性,因此偏振成像技术可适用于多种复杂环境中的智能监控、交通监察领域。近年来使用深度学习判读图像检测目标的方法迅速发展,已经广泛应用于图像处理的各个领... 详细信息

偏振光反射信息可直接反演目标本征特性,且在传输过程中具备较强的抗干扰特性,因此偏振成像技术可适用于多种复杂环境中的智能监控、交通监察领域。近年来使用深度学习判读图像检测目标的方法迅速发展,已经广泛应用于图像处理的各个领域。本文提出了一种基于偏振图像与深度神经网络算法的行人、车辆多目标检测算法YOLOv5s-DOLP。首先,通过实时获取到偏振图像进行偏振信息解析,获取目标偏振度图像。其次,为增强偏振度图像中检测目标与背景存在高对比度的特性,在主干网络中引入通道注意力与空间注意力,提升网络特征进行自适应学习的能力。此外,使用K-means算法对目标位置信息进行聚类分析,加快网络在偏振度图像的学习速度,提升目标检测精度。实验结果显示,该算法结合了偏振成像和深度学习目标检测的优势,对于低照度复杂场景中的车辆、行人目标检测效果好、检测速度快,对于道路车辆的目标检测、识别与跟踪具有一定的应用价值。

关键词：偏振成像神经网络 YOLOv5s 多目标检测注意力机制

来源：评论

学校读者我要写书评

暂无评论

PLGS: Robust Panoptic Lifting with 3D Gaussian Splatting

引用

IEEE transactions on image processing : a publication of the IEEE signal processing Society 2025年 PP卷 PP页

作者： Yu Wang Xiaobao Wei Ming Lu Guoliang Kang School of Automation Science and Electrical Engineering Beihang University Beijing China Institute of Software Chinese Academy of Sciences Beijing China Institute of Software University of Chinese Academy of Sciences Beijing China School of Computer Science Peking University Beijing China

Previous methods utilize the neural Radiance Field (NeRF) for panoptic lifting, while their training and rendering speed are unsatisfactory. In contrast, 3D Gaussian Splatting (3DGS) has emerged as a prominent technique due to its rapid training and rendering speed. However, unlike NeRF, the conventional 3DGS may not satisfy the basic smoothness assumption as it does not rely on any parameterized structures to render (e.g., MLPs). Consequently, the conventional 3DGS is, in nature, more susceptible to noisy 2D mask supervision. In this paper, we propose a new method called PLGS that enables 3DGS to generate consistent panoptic segmentation masks from noisy 2D segmentation masks while maintaining superior efficiency compared to NeRF-based methods. Specifically, we build a panoptic-aware structured 3D Gaussian model to introduce smoothness and design effective noise reduction strategies. For the semantic field, instead of initialization with structure from motion, we construct reliable semantic anchor points to initialize the 3D Gaussians. We then use these anchor points as smooth regularization during training. Additionally, we present a self-training approach using pseudo labels generated by merging the rendered masks with the noisy masks to enhance the robustness of PLGS. For the instance field, we project the 2D instance masks into 3D space and match them with oriented bounding boxes to generate cross-view consistent instance masks for supervision. Experiments on various benchmarks demonstrate that our method outperforms previous state-of-the-art methods in terms of both segmentation quality and speed.

关键词： Three-dimensional displays Semantics Training Rendering (computer graphics) Noise measurement neural radiance field Solid modeling image reconstruction Point cloud compression Noise

来源：评论

学校读者我要写书评

暂无评论

Pan-sharpening via a gradient-based deep network prior

引用

signal processing-image COMMUNICATION 2019年 74卷 322-331页

作者： Ye, Fei Guo, Yecai Zhuang, Peixian Nanjing Univ Informat Sci & Technol Sch Elect & Informat Engn Nanjing 210044 Jiangsu Peoples R China

Pan-sharpening is a domain-specific task of satellite imagery fusion. However, most traditional methods fuse the panchromatic image and the multispectral images in linear manners, which lead to severe spectral and spatial distortions. In the meanwhile, discriminative learning methods are limited in specialized satellites and tasks. In this paper, we make an attempt to integrate a deep prior with model-based optimization scheme for pan-sharpening. The proposed deep prior is based on a convolutional neural network which is composed of the proposed problem-specific recursive block and is trained in gradient domain. We plug the trained prior in place of the spatial preservation term in model-based optimization scheme, and address it with the alternating direction method of multipliers. Final experimental results demonstrate that the proposed model can overcome the restriction of linear model, and greatly reduce spectral and spatial distortions. Compared with several discriminative learning methods, our model tends to achieve promising generalization across different satellites.

关键词： Pan-sharpening Model-based optimization Convolutional neural network Gradient-based prior

来源：评论

学校读者我要写书评

暂无评论

Multi-frequency point source detection with fully convolutional networks: Performance in realistic microwave sky simulations

引用

ASTRONOMY & ASTROPHYSICS 2022年 658卷

作者： Casas, J. M. Gonzalez-Nuevo, J. Bonavera, L. Herranz, D. Suarez Gomez, S. L. Cueli, M. M. Crespo, D. Santos, J. D. Sanchez, M. L. Sanchez-Lasheras, F. de Cos, F. J. Univ Oviedo Dept Fis C Federico Garcia Lorca 18 Oviedo 33007 Spain Inst Univ Ciencias & Tecnol Espaciales Asturias I C Independencia 13 Oviedo 33004 Spain Univ Oviedo Dept Matemat C Federico Garcia Lorca 18 Oviedo 33007 Spain Univ Cantabria Dept Fis Moderna Santander 39005 Spain CSIC UC Inst Cantabria Av Los Castros S-N Santander 39005 Spain Escuela Ingn Minas Energia & Mat Independencia 13 Oviedo 33004 Spain

Context. Point source (PS) detection is an important issue for future cosmic microwave background (CMB) experiments since they are one of the main contaminants to the recovery of CMB signal on small scales. Improving its multi-frequency detection would allow us to take into account valuable information otherwise neglected when extracting PS using a channel-by-channel approach. Aims. We aim to develop an artificial intelligence method based on fully convolutional neural networks to detect PS in multi-frequency realistic simulations and compare its performance against one of the most popular multi-frequency PS detection methods, the matrix filters. The frequencies used in our analysis are 143, 217, and 353 GHz, and we imposed a Galactic cut of 30 degrees. methods. We produced multi-frequency realistic simulations of the sky by adding contaminating signals to the PS maps as the CMB, the cosmic infrared background, the Galactic thermal emission, the thermal Sunyaev-Zel'dovich effect, and the instrumental and PS shot noises. These simulations were used to train two neural networks called flat and spectral MultiPoSeIDoNs. The first one considers PS with a flat spectrum, and the second one is more realistic and general because it takes into account the spectral behaviour of the PS. Then, we compared the performance on reliability, completeness, and flux density estimation accuracy for both MultiPoSeIDoNs and the matrix filters. Results. Using a flux detection limit of 60 mJy, MultiPoSeIDoN successfully recovered PS reaching the 90% completeness level at 58 mJy for the flat case, and at 79, 71, and 60 mJy for the spectral case at 143, 217, and 353 GHz, respectively. The matrix filters reach the 90% completeness level at 84, 79, and 123 mJy. To reduce the number of spurious sources, we used a safer 4 sigma flux density detection limit for the matrix filters, the same as was used in the Planck catalogues, obtaining the 90% of completeness level at 113, 92, and 398 mJy. In all

关键词： techniques image processing submillimeter galaxies cosmic background radiation

来源：评论

学校读者我要写书评

暂无评论

Distributed signal Strength Prediction using Satellite Map empowered by Deep Vision Transformer

Distributed Signal Strength Prediction using Satellite Map e...

引用

IEEE Globecom Workshops

作者： Haiyao Yu Zhanwei Hou Yifan Gu Peng Cheng Wanli Ouyang Yonghui Li Branka Vucetic School of Electrical and Information Engineering University of Sydney Sydney NSW Australia Computer Science & Information Technology La Trobe University Melbourne VIC Australia

ISBN: (纸本)9781665423915

The accurate prediction of received signal strength (RSS) is the key to coverage optimization and interference management in network planning, as well as proactive resource allocation and anticipated network management. Traditional methods for RSS prediction are based on ray tracing or stochastic radio propagation model. The former requires the detailed 3D geometry and dielectric properties of the reflectors, which may not be available practically. The latter roughly classify the environment as either urban, suburban and rural scenarios and does not make full use of the environment information. In this paper, by leveraging accessible satellite maps to capture the features of radio environment, a distributed federated learning (FL) RSS prediction framework is proposed to fully exploit the user generated real-time data while preserving the users’ privacy. To further improve the prediction accuracy, the deep vision transformer (DeepVIT) is utilized to process the images of the satellite map, because it is capable of learning to "pay attention to" important parts of an image such as reflection surfaces and blockages. The proposed method is evaluated by the real-world data set including around 60, 000 individual measurements. Simulations results verified that the prediction accuracy of the proposed method outperforms baseline methods including ray tracing, Urban Macro (UMa) model and convolutional neural network (CNN) based method. Moreover, the computational time is reduced five times compared with CNN based method.

关键词： Computational modeling Satellite broadcasting Distributed databases stochastic processes Ray tracing Radio propagation Predictive models

来源：评论

学校读者我要写书评

暂无评论

3-D super-resolution localization microscopy using deep learning method 9

3-D super-resolution localization microscopy using deep lear...

引用

Conference on Optics in Health Care and Biomedical Optics IX held at SPIE/COS Photonics Asia Conference

作者： Lu, Mengyang Zhou, Tianyang Liu, Xin Shanghai Univ Sch Commun & Informat Engn Shanghai 200444 Peoples R China

ISBN: (纸本)9781510630987

Super-resolution localization microscopy (SRLM) techniques overcome the diffraction limit, making possible the observation of sub-cellular structures in vivo. At present, the spatial resolution of similar to 20 nm in x-y axis has been achieved in SRLM. However, the localization accuracy for the longitudinal axis (i.e., the z-axis) still need be improved. Although some methods have been proposed to implement 3-D SRLM, these methods are computationally intensive and parameter dependent. To overcome these limitations, in this paper, we propose a new method based on deep learning, termed as dl-3D-SRLM. By learning the mapping between a 2-D camera frame (i.e., the experimentally acquired image) and the true 3-D locations of fluorophores in the corresponding image region with a convolutional neural network (CNN), dl-3D-SRLM provides the possibility of implementing 3-D SRLM with a high localization accuracy, a fast data-processing speed, and a little human intervention. To evaluate the performance of dl-3D-SRLM, a series of numerical simulations are performed. The results show that when using dl-3D-SRLM, we can accurately resolve the 3-D location of fluorophores from the acquired 2-D images, even if under high fluorophores densities and low signal-to-noise ratio conditions. In addition, the complex 3-D structure can also be effectively imaged by dl-3D-SRLM. As a result, dl-3D-SRLM is more beneficial for 3D-SRLM imaging.

关键词： Super-resolution localization microscopy convolutional neural networks imaging reconstruction deep learning 3-D imaging

来源：评论

学校读者我要写书评

暂无评论

neural network based corn field furrow detection for autonomous navigation in agriculture vehicles

Neural network based corn field furrow detection for autonom...

引用

2020 IEEE International IOT, Electronics and Mechatronics Conference, IEMTRONICS 2020

作者： Anthony Simon, Niko Min, Cheol-Hong University of St. Thomas Department of Electrical Engineering St. PaulMN United States

ISBN: (纸本)9781728196152

Row detection in agricultural applications has commonly used Hough transform techniques and traditional signal processing based approaches relating to machine vision. There are various learning based methods available that are capable of producing similar results in terms of detection. In this paper, a neural network based algorithm is developed, and we compare the Hough transform and a machine learning implementation with the proposed approach to determine which would be the most appropriate in a real-time application given a variety of factors including computational performance, accuracy, and environmental variability. Compared to the learning based approaches which rely on training data, Hough transform based detection relies on a variety of processes, including binarization and denoising, which are not required to be explicitly implemented in the machine learning or neural network models. Additionally, to add another layer of diversity to the three possible solutions examined is the consideration for color input data. The Hough transform method and the neural network model implemented both require color input data while the machine learning model relies on texture features instead of color to make its classification predictions. Compared to the traditional image understanding techniques, autonomous vehicles face challenges due to similarities in color and texture between the crops and their surroundings. Therefore, the algorithm is developed to overcome such challenges. Preliminary results show that the neural network model developed was found to offer the most versatility compared to traditional methods and the highest accuracy on the order of 97% for this application across several different input conditions. © 2020 IEEE.

关键词： Computer vision

来源：评论

学校读者我要写书评

暂无评论

Wavelet Deep neural Network for Stripe Noise Removal

引用

IEEE ACCESS 2019年 7卷 44544-44554页

作者： Guan, Juntao Lai, Rui Xiong, Ai Xidian Univ Dept Microelect Xian 710071 Shaanxi Peoples R China Chengdu Univ Informat Technol Sch Control Engn Chengdu 610103 Sichuan Peoples R China

The stripe noise effects severely degrade the image quality in infrared imaging systems. The existing destriping algorithms still struggle to balance noise suppression, detail preservation, and real-time performance, which retards their application in spectral imaging and signal processing field. To solve this problem, an innovative wavelet deep neural network from the perspective of transform domain is presented in this paper, which takes the intrinsic characteristics of stripe noise and complementary information between the coefficients of different wavelet sub-bands into full consideration to accurately estimate the noise with the lower computational load. In addition, a special directional regularizer is further defined to separate the scene details from stripe noise more thoroughly and recover the details more accurately. The extensive experiments on simulated and real data demonstrate that our proposed method outperforms several classical destriping methods on both quantitative and qualitative assessments.

关键词： neural networks image denoising infrared image sensors wavelet transforms

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：