检索结果-内蒙古大学图书馆

FISS GAN:A Generative Adversarial Network for Foggy Image Semantic Segmentation

IEEE/CAA Journal of Automatica Sinica 2021年第8期8卷 1428-1439页

作者： Kunhua Liu Zihao Ye Hongyan Guo Dongpu Cao Long Chen Fei-Yue Wang School of Computer Science and Engineering Sun Yat-sen UniversityGuangzhou 510006China State Key Laboratory of Automotive Simulation and Control and the Department of Control Science and Engineering Jilin University(Campus Nanling)Changchun 130025China Mechanical and Mechatronics Engineering Department at the University of Waterloo 200 University Avenue West WaterlooON N2L 3G1Canada State Key Laboratory of Management and Control for Complex Systems Institute of AutomationChinese Academy of SciencesBeijing 100190 Institute of Systems Engineering Macao University of Science and TechnologyMacao 999078China

Because pixel values of foggy images are irregularly higher than those of images captured in normal weather(clear images),it is difficult to extract and express their *** method has previously been developed to directly explore the relationship between foggy images and semantic segmentation *** investigated this relationship and propose a generative adversarial network(GAN)for foggy image semantic segmentation(FISS GAN),which contains two parts:an edge GAN and a semantic segmentation *** edge GAN is designed to generate edge information from foggy images to provide auxiliary information to the semantic segmentation *** semantic segmentation GAN is designed to extract and express the texture of foggy images and generate semantic segmentation *** on foggy cityscapes datasets and foggy driving datasets indicated that FISS GAN achieved state-of-the-art performance.

关键词： Edge GAN foggy images foggy image semantic segmentation GAN semantic segmentation

来源：评论

学校读者我要写书评

暂无评论

Biomechanical models in the lower-limb exoskeletons development: A review

TechRxiv

引用

TechRxiv 2023年

作者： Firouzi, Vahid Seyfarth, Andre Song, Seungmoon Von Stryk, Oskar Sharbafi, Maziar Ahmad Lauflabor Locomotion Laboratory Institute of Sport Science TU Darmstadt Germany Simulation Systems Optimization and Robotics Group Department of Computer Science TU Darmstadt Germany Mechanical and Industrial Engineering Northeastern University United States

Lower limb exoskeletons serve multiple purposes, like supporting and augmenting movement. Biomechanical models are practical tools to understand human movement, and motor control. This review explores the applications of biomechanical gait models in the development of lower-limb exoskeletons. Biomechanical models can be broadly categorized as conceptual and detailed models. Biomechanical models can be used for the design, control, and assessment of exoskeletons. Also, these models can be used to estimate unmeasurable or hard-to-measure variables, which is also useful within the aforementioned applications. The analysis of applications and research areas aims to inspire further advancements in this field and contribute to the development of more effective exoskeletons. We highlight validating simulation studies, improving the accuracy and fidelity of biomechanical models, and benefiting from exoskeletons to enhance or validate biomechanical models as future research areas related to biomechanical models. We also emphasize the exploration of model-based design and control approaches for exoskeletons targeting pathological gait, and utilizing biomechanical models for diverse design objectives of exoskeletons. Also, increasing the availability of open-source resources accelerates exoskeleton and biomechanical model advancement. Our review reveals existing needs and cracks new perspectives for the development of more usable and human-compatible exoskeletons based on biomechanical models. © 2023, CC BY.

关键词： Exoskeleton (Robotics)

来源：评论

学校读者我要写书评

暂无评论

Detection of Detonators in Baggage Using Deep CNN Architecture: A Comparative Study of Contrast Enhancement and Feature Extraction Methods

Detection of Detonators in Baggage Using Deep CNN Architectu...

引用

International Conference on Image and Signal Processing and their Applications (ISPA)

作者： Oulhissane Lynda Merah Mostefa Simona Moldovanu Luminita Moraru Laboratory of Signals and Systems LSS Faculty of Science and Technology Abdelhamid Ibn Badis University Mostaganem Algeria MSlab Department of Computer Science and Information Technology Modelling & Simulation Laboratory Faculty of Automation Computers Electrical Engineering and Electronics Dunarea de Jos University of Galati Galati Romania MSlab Department of Chemistry Physics & Environment Modelling & Simulation Laboratory Faculty of Sciences and Environment Dunarea de Jos University of Galati Galati Romania

ISBN: (数字)9798350309249

ISBN: (纸本)9798350309256

Visual inspection of dual-energy X-ray radiographic images of cabin baggage requires high performance, but is hindered by various challenges such as low target prevalence, variability in target visibility, possible presence of multiple targets, and security personnel fatigue and inattention. Artificial intelligence (AI) techniques, particularly deep convolutional neural networks (CNNs), have shown promise in improving the automatic detection of explosives, even with low-resolution radiographic images, especially in high baggage throughput scenarios. In this paper, we focus on the detection of detonators as components of improvised explosive devices. The proposed approach involves comparing two experiments implemented in a deep CNN architecture using TensorFlow and Keras libraries. In the first experiment, raw dual-energy radiographic images without any enhancement were used. The second experiment includes three methods for contrast enhancement and feature extraction: the Contrast Limited Adaptive Histogram Equalization (CLAHE) method, the wavelet transform-based method, and the mixed CLAHE RGB-Wavelet method. In the latter two methods, Haar, Db2, Coif2, and Sym2 mother wavelet functions at two levels (HH and HL) were employed. The analysis of results focuses on a comparative study of performance measures such as accuracy, precision, recall, and F1 score. It was found that the preprocessing methods used in experiment 2, for the two evaluated classes (detonator and no detonator), achieved higher accuracy compared to the raw radiographic images used in experiment 1 (98.08%). The highest accuracies in experiment 2, with a value of 100%, were obtained with the CLAHE method (green channel in grayscale, blue channel in grayscale, and RGB channels) and the wavelet transform method with Haar mother wavelet at two levels HL

关键词： Wavelet transforms Radiography Visualization Gray-scale Inspection Feature extraction Libraries

来源：评论

学校读者我要写书评

暂无评论

Melanoma Detection using a Random Forest Algorithm 10

Melanoma Detection using a Random Forest Algorithm

引用

10th E-Health and Bioengineering Conference, EHB 2022

作者： Damian, Felicia-Anisoara Moldovanu, Simona Moraru, Luminita The Modelling & Simulation Laboratory University Dunarea de Jos of Galati UGAL Galati Romania Department of Computer Science and Information Technology Faculty of Automation Computers Electrical Engineering and Electronics University Dunarea de Jos of Galati UGAL Galati Romania Department of Chemistry Physics and Environment Faculty of Sciences and Environment University Dunarea de Jos of Galati UGAL Galati Romania

ISBN: (纸本)9781665485579

In recent years, statistical evidence has identified melanoma as a major cause of skin cancer deaths. In this study, we propose a skin analysis lesions based on colour features (C) and Hu (H) moments to differentiate benign (nevi) and malignant (melanoma) lesions and, a classification based on Random Forest algorithm. Colour features are found as follows. Pre-processing for hair removing using the Dull-razor method;image segmentation based on the optimal colour threshold set on each RGB colour channel (red, green, blue);23 colour features extraction using the minimum and maximum RGB intensity and Hu moments;Random Forest (RF) algorithm to accurately differentiate benign and malignant skin lesions. This study used three datasets: 7-Point, Med-Node, PH2. RF classifier performance in terms of accuracy was: 94% for the 7-point dataset with H5, H7, H6, H2, C08 selected as relevant features;98.1% for the Med-Node dataset and H7, H5, H6, H2, H1 ACC the relevant features;and, 96.5% for the PH2 dataset and H5, H6, H7, C03, H1 selected as relevant features, respectively. Shape features based on Hu moments are found to improve the classification performance of the skin lesions compared to the colour features. © 2022 IEEE.

关键词： Image segmentation

来源：评论

学校读者我要写书评

暂无评论

Well-posedness analysis of the Cahn-Hilliard-Biot model

arXiv

引用

arXiv 2023年

作者： Riethmüller, Cedric Storvik, Erlend Both, Jakub Wiktor Radu, Florin Adrian Institute of Applied Analysis and Numerical Simulation University of Stuttgart Germany Department of Computer Science Electrical Engineering and Mathematical Sciences Western Norway University of Applied Sciences Norway Center for Modeling of Coupled Subsurface Dynamics Department of Mathematics University of Bergen Norway

We investigate the well-posedness of the recently proposed Cahn-Hilliard-Biot model. The model is a three-way coupled PDE of elliptic-parabolic nature, with several nonlinearities and the fourth order term known to the Cahn-Hilliard system. We show existence of weak solutions to the variational form of the equations and uniqueness under certain conditions of the material parameters and secondary consolidation, adding regularizing effects. Existence is shown by discretizing in space and applying ODE-theory (the Peano-Cauchy theorem) to prove existence of the discrete system, followed by compactness arguments to retain solutions of the continuous system. In addition, the continuous dependence of solutions on the data is established, in particular implying uniqueness. Both results build strongly on the inherent gradient flow structure of the model. Copyright © 2023, The Authors. All rights reserved.

关键词： Variational techniques

来源：评论

学校读者我要写书评

暂无评论

Efficient Training Approaches for Performance Anomaly Detection Models in Edge Computing Environments

arXiv

引用

arXiv 2024年

作者： Fernando, Duneesha Rodriguez, Maria A. Arroba, Patricia Ismail, Leila Buyya, Rajkumar Laboratory School of Computing and Information Systems The University of Melbourne Parkville Australia The CCS Center for Computational Simulation Universidad Politécnica de Madrid Madrid Spain Lab Department of Computer Science and Software Engineering United Arab Emirates University United Arab Emirates

Microservice architectures are increasingly used to modularize IoT applications and deploy them in distributed and heterogeneous edge computing environments. Over time, these microservice-based IoT applications are susceptible to performance anomalies caused by resource hogging (e.g., CPU or memory), resource contention, etc., which can negatively impact their Quality of Service and violate their Service Level Agreements. Existing research on performance anomaly detection for edge computing environments focuses on model training approaches that either achieve high accuracy at the expense of a time-consuming and resource-intensive training process or prioritize training efficiency at the cost of lower accuracy. To address this gap, while considering the resource constraints and the large number of devices in modern edge platforms, we propose two clustering-based model training approaches: (1) intra-cluster parameter transfer learning-based model training (ICPTL) and (2) cluster-level model training (CM). These approaches aim to find a trade-off between the training efficiency of anomaly detection models and their accuracy. We compared the models trained under ICPTL and CM to models trained for specific devices (most accurate, least efficient) and a single general model trained for all devices (least accurate, most efficient). Our findings show that ICPTL's model accuracy is comparable to that of the model per device approach while requiring only 40% of the training time. In addition, CM further improves training efficiency by requiring 23% less training time and reducing the number of trained models by approximately 66% compared to ICPTL, yet achieving a higher accuracy than a single general model. Copyright © 2024, The Authors. All rights reserved.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Improved Classification of Benign and Malignant Breast Lesions Based on Contrast Manipulation and Gradient Map 10

Improved Classification of Benign and Malignant Breast Lesio...

引用

10th E-Health and Bioengineering Conference, EHB 2022

作者： Nastase, Iulia-Nela Anghelache Moldovanu, Simona Moraru, Luminita The Modelling & Simulation Laboratory University Dunarea de Jos of Galati UGAL Galati Romania Department of Computer Science and Information Technology Faculty of Automation Computers Electrical Engineering and Electronics University Dunarea de Jos of Galati UGAL Galati Romania Department of Chemistry Physics and Environment Faculty of Sciences and Environment University Dunarea de Jos of Galati UGAL Galati Romania

ISBN: (纸本)9781665485579

This study proposes an approach to enhance the contrast breast ultrasound (BUS) images with the preservation of the finer details. The main idea is to use the relevant and most important features in the classification of BUS images as benign or malignant. Among the contrast manipulation techniques, both linear under- and overexposure methods, contrast limiting adaptive histogram equalization (CLAHE) method and homomorphic filtering are used. The gradient maps of original images are generated for further analysis. Also, the gradient maps of original images are improved by the local mean subtraction and divisive normalization. Eight structural features are extracted from each gradient map image type. The extracted features feed a Random Forest (RF) classifier to determine their relevance for BUS images classification. An accuracy of 97.12% has been obtained for improved gradient map images and 93.03% for gradient map images. © 2022 IEEE.

关键词： Image classification

来源：评论

学校读者我要写书评

暂无评论

Retrieving Soil Organic Matter and Soil Moisture Profiles of the Arctic Foothills Tundra Using P-band Polarimetric SAR Imagery

Retrieving Soil Organic Matter and Soil Moisture Profiles of...

引用

IEEE International Symposium on Geoscience and Remote Sensing (IGARSS)

作者： Kazem Bakian-Dogaheh Yuhuan Zhao John S. Kimball Mahta Moghaddam Ming Hsieh Department of Electrical and Computer Engineering University of Southern California Los Angeles CA USA Numerical Terradynamic Simulation Group W. A. Franke College of Forestry & Conservation University of Montana Missoula MT

This paper presents a physics-based radar modeling framework that enables joint retrievals of the Arctic tundra permafrost active layer soil organic matter content and soil moisture profile using P-band polarimetric SAR. Initially, an extensive set of field observations are used to model the subsurface soil moisture and organic matter profiles with independent model parameters. Then a new organic soil dielectric model is used to translate the soil profile properties into equivalent dielectric properties that bridge the soil field properties to their manifestation in radar measurements. Finally, a multi-layered dielectric structure is adopted by an electromagnetic scattering computational model that predicts equivalent backscattering coefficients resulting from the soil profile state parameters. These pieces together construct the forward model, which is at the heart of a physics-based radar retrieval algorithm. Finally, we integrate the developed model into a retrieval scheme, in which the soil moisture and organic profile model parameters are estimated using radar backscattering coefficients measured by AirMOSS P-band radar. We show retrieved soil moisture and soil organic matter profiles, derived pixel-wise by the algorithm providing the first airborne-driven soil organic carbon map.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Iris-Based Biometric Identification Using a Combination of the Right - Left Iris Statistical Features 12

Iris-Based Biometric Identification Using a Combination of t...

引用

12th International Conference on Mathematical Modeling in Physical Sciences, IC-MSQUARE 2023

作者： Dincǎ Lǎzǎrescu, A.M. Moldovanu, S. Moraru, L. The Modelling & Simulation Laboratory Dunarea de Jos University of Galati 47 Domneasca Street Galati800008 Romania Mihail Kogǎlniceanu High School 161B Brǎilei St. Galati800320 Romania Department of Computer Science and Information Technology Faculty of Automation Computers Electrical Engineering and Electronics Dunarea de Jos University of Galati 47 Domneasca Street Galati800008 Romania Department of Chemistry Physics & Environment Faculty of Sciences and Environment Dunarea de Jos University of Galati 47 Domneasca Street Galati800008 Romania

A combination between the information extracted for both right iris and left iris could increase the efficacy of the biometric recognition systems. In this paper, we propose a biometric identification method based on density of image patterns extracted from human iris images and the combination and comparison of the right iris and the left iris characteristics. The density of the patters approach for processed images can be a new biometric feature used to implement a biometric recognition system with high performance when a small feature dimension is used. In this way, we can maximize the retention of the effective biometric information. The experiments were conducted on the MMU Iris Database containing 225 images of the left eye and 225 images of the right eye. Two morphological Top-hat and Hit or Miss transforms were implemented to find out the particular pattern of pixels. They allow for the enhancement of detail in images. Then, a statistical feature extraction technique is employed to derive the density of the patterns in morphological transformed images. To assess the density of the patterns differences between the right and left iris data groups, the Pearson's correlation coefficient (PCC) is computed. We reported very good results with a PCC of 0.6164 (strong and positive correlation) for Top-hat morphological operation whilst the Hit or Miss transform returns a PCC of 0.0127 so there is no relationship between the density of the patterns in the right and left irises. © 2024 Institute of Physics Publishing. All rights reserved.

关键词： Biometrics

来源：评论

学校读者我要写书评

暂无评论

3S-TSE: EFFICIENT THREE-STAGE TARGET SPEAKER EXTRACTION FOR REAL-TIME AND LOW-RESOURCE APPLICATIONS

arXiv

引用

arXiv 2023年

作者： He, Shulin Liu, Jinjiang Li, Hao Yang, Yang Chen, Fei Zhang, Xueliang College of Computer Science Inner Mongolia University China Department of Electrical and Electronic Engineering Southern University of Science and Technology Shenzhen China Key Laboratory of Urban Land Resources Monitoring and Simulation Ministry of Natural Resources China

Target speaker extraction (TSE) aims to isolate a specific voice from multiple mixed speakers relying on a registerd sample. Since voiceprint features usually vary greatly, current end-to-end neural networks require large model parameters which are computational intensive and impractical for real-time applications, espetially on resource-constrained platforms. In this paper, we address the TSE task using microphone array and introduce a novel three-stage solution that systematically decouples the process: First, a neural network is trained to estimate the direction of the target speaker. Second, with the direction determined, the Generalized Sidelobe Canceller (GSC) is used to extract the target speech. Third, an Inplace Convolutional Recurrent Neural Network (ICRN) acts as a denoising post-processor, refining the GSC output to yield the final separated speech. Our approach delivers superior performance while drastically reducing computational load, setting a new standard for efficient real-time target speaker extraction. Copyright © 2023, The Authors. All rights reserved.

关键词： Extraction

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：