检索结果-内蒙古大学图书馆

Anomaly-Focused Augmentation Method for Industrial Visual Inspection

IEEE Access 2025年 13卷 89118-89139页

作者： Melvinlara, Melvin A. Hameed, Ibrahim A. Koziorek, Jiri Faculty of Electrical Engineering and Computer Science Department of Cybernetics and Biomedical Engineering Ostrava70800 Czech Republic Department of Mechanical Engineering and Technology Management Ås1433 Norway

In industrial inspection, the detection of surface defects - such as scratches, dents, or other defects - is crucial for ensuring product quality. However, the limited availability of annotated images of such defects poses challenges for developing reliable detection models. This paper proposes a novel image augmentation algorithm to address this limitation by generating synthetic imperfections from a small sample set. Unlike traditional approaches that augment entire images, the proposed approach isolates defects and applies spatial transformations (e.g., rotation, shear) before blending them into background images. This allows for the creation of diverse training datasets. The algorithm employs a 2D binary map to encode the spatial relationship between defect and background regions of images and then applies an AI-driven model to generate new realistic blended images. By leveraging a generator-discriminator framework, the augmentation process is iteratively refined during training to maintain the style and texture of the original imperfections. Experimental results show that the proposed method significantly enhances the performance of defect detection models compared to existing augmentation techniques while preserving the realism of the generated images. © 2013 IEEE.

关键词： Photointerpretation

来源：评论

学校读者我要写书评

暂无评论

Deep Learning Approach for the Reconstruction of Droplet Spreading Based on Wettability 1

Deep Learning Approach for the Reconstruction of Droplet Spr...

引用

1st International Conference on AIML-Applications for engineering and Technology, ICAET 2025

作者： Meshram, Ganesh Sahadeo Chakraborty, Suman Chakrabarti, Partha P Mechanical Engineering IIT Kharagpur Kharagpur India Computer Science & Engineering IIT Kharagpur Kharagpur India

ISBN: (纸本)9798350355611

Convolutional autoencoders are used in this research to effectively reconstruct the droplet spreading on smooth surfaces with different wettabilities based on contact angles. Accurate modeling of water droplets on smooth surfaces is achieved using the lattice Boltzmann method (LBM), which later provided data for the CNN-AE to be trained on. Before reconstructing the wettability data by upsampling and ReLU activation, the CNN-AE compresses the data into a lower-dimensional latent space by using convolutional layers and max pooling. The MSE and the structural similarity index were used to evaluate the huge wettability dataset produced by LBM simulations centered around random barriers. Wetability fields are efficiently reconstructed by the model, which has 112,579 trainable parameters. Within the realm of machine learning-based microfluidics research, this work reveals that CNN-AE is capable of properly reconstructing complicated droplet spreading patterns, which paves the way for further developments, including textured surfaces. © 2025 IEEE.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Noise4Denoise:Leveraging noise for unsupervised point cloud denoising

引用

Computational Visual Media 2024年第4期10卷 659-669页

作者： Weijia Wang Xiao Liu Hailing Zhou Lei Wei Zhigang Deng Manzur Murshed Xuequan Lu Deakin University GeelongVIC 3216Australia Department of Mechanical Engineering and Product Design Engineering Swinburne University of TechnologyMelbourneVIC 3122Australia Department of Computer Science University of HoustonHoustonTX 77204-3027USA Department of Computer Science and Information Technology La Trobe UniversityMelbourneVIC 3086Australia

Existing deep learning-based point cloud denoising methods are generally trained in a supervised manner that requires clean data as ground-truth ***,in practice,it is not always feasible to obtain clean point *** this paper,we introduce a novel unsupervised point cloud denoising method that eliminates the need to use clean point clouds as groundtruth labels during *** demonstrate that it is feasible for neural networks to only take noisy point clouds as input,and learn to approximate and restore their clean *** particular,we generate two noise levels for the original point clouds,requiring the second noise level to be twice the amount of the first noise *** this,we can deduce the relationship between the displacement information that recovers the clean surfaces across the two levels of noise,and thus learn the displacement of each noisy point in order to recover the corresponding clean *** experiments demonstrate that our method achieves outstanding denoising results across various datasets with synthetic and real-world noise,obtaining better performance than previous unsupervised methods and competitive performance to current supervised methods.

关键词： point clouds denoising geometry processing deep learning

来源：评论

学校读者我要写书评

暂无评论

Temperature-triggered liquid metal actuators for fluid manipulation by leveraging phase transition control

引用

International Journal of Smart and Nano Materials 2024年第4期15卷 730-742页

作者： Hongda Lu Jiayi Yang Mengqing Zhao Qingtian Zhang Jialu Wang Xiangbo Zhou Yipu Guo Liping Gong Zexin Chen Shi-Yang Tang Weihua Li School of Mechanical MaterialsMechatronic and Biomedical EngineeringUniversity of WollongongWollongongAustralia School of Mechanical Electronic and Control EngineeringBeijing Jiaotong UniversityBeijingP.R.China School of Mechanical and Manufacturing Engineering University of New South WalesSydneyAustralia School of Electronics and Computer Science University of SouthamptonSouthamptonUK

Small-scale pumps for controlling microfluidics have promising applications in drug delivery and chemical *** metal(LM)demonstrates excellent flow pumping performance due to its simple structure and the electrocapillary effect under an electric ***,LM droplets risk escaping from constrained structures,which can lead to pump *** regulation is also a critical parameter in optimizing chemical reactions in fluidic systems,however,integrating it into a compact system remains ***,we develop a temperature-triggered gallium-based actuator(TTGA)by introducing a gallium(Ga)droplet wetted on a copper(Cu)plate as the core element for flow *** Cu plate prevents the Ga droplet from escaping the chamber and significantly increases the flow *** leveraging the electrochemical method to inhibit the supercooling effect of Ga,the TTGA enables activation/deactivation for flow actuation at different *** investigate the impact of electrode position,solution concentration,and applied voltage on TTGA’s pumping *** dynamically tuning the Ga droplet’s temperature to control phase transition,TTGA allows for accurate flow actuation ***,placing Ga and eutectic Ga-indium(EGaIn)droplets in different channels enables the expected flow divergence for fluids with different *** development of TTGA presents new opportunities in microfluidics and biomedical treatment.

关键词： Liquid metal actuators temperature sensation phase transition

来源：评论

学校读者我要写书评

暂无评论

Heterogeneous Task Allocation Model and Algorithm for Intelligent Connected Vehicles

引用

computers, Materials & Continua 2024年第9期80卷 4281-4302页

作者： Neng Wan Guangping Zeng Xianwei Zhou School of Computer and Communication Engineering University of Science and Technology BeijingBeijing100083China School of Mechanical and Transportation Southwest Forestry UniversityKunming650224China

With the development of vehicles towards intelligence and connectivity,vehicular data is diversifying and growing dramatically.A task allocation model and algorithm for heterogeneous Intelligent Connected Vehicle(ICV)applications are proposed for the dispersed computing network composed of heterogeneous task vehicles and Network Computing Points(NCPs).Considering the amount of task data and the idle resources of NCPs,a computing resource scheduling model for NCPs is *** the heterogeneous task execution delay threshold as a constraint,the optimization problem is described as the problem of maximizing the utilization of computing resources by *** proposed problem is proven to be NP-hard by using the method of reduction to a 0-1 knapsack problem.A many-to-many matching algorithm based on resource preferences is *** algorithm first establishes the mutual preference lists based on the adaptability of the task requirements and the resources provided by *** enables the filtering out of un-schedulable NCPs in the initial stage of matching,reducing the solution space *** solve the matching problem between ICVs and NCPs,a new manyto-many matching algorithm is proposed to obtain a unique and stable optimal matching *** simulation results demonstrate that the proposed scheme can improve the resource utilization of NCPs by an average of 9.6%compared to the reference scheme,and the total performance can be improved by up to 15.9%.

关键词： Task allocation intelligent connected vehicles dispersed computing matching algorithm

来源：评论

学校读者我要写书评

暂无评论

Parameter Efficient Point Cloud Prompt Tuning for Unified Point Cloud Understanding

IEEE Transactions on Intelligent Vehicles

引用

IEEE Transactions on Intelligent Vehicles 2024年 1-16页

作者： Fei, Ben Liu, Liwen Yang, Weidong Li, Zhijun Chen, Wen-Ming Ma, Lipeng School of Computer Science Fudan University Shanghai China School of Mechanical Engineering Tongji University Shanghai China Academy for Engineering and Technology Fudan University Shanghai China

The prevailing paradigm in 3D vision involves fully fine-tuning all the backbone parameters of pre-trained models. However, this approach poses challenges due to the large number of parameters requiring tuning, resulting in unexpected storage demands. To address these issues and alleviate the computational cost and storage burden associated with full fine-tuning, we propose Point Cloud Prompt Tuning (PCPT) as an effective method for large Transformer models in point cloud processing. PCPT offers a powerful and efficient solution to mitigate the costs associated with full fine-tuning. Drawing inspiration from recent advancements in efficient tuning of large-scale language models and 2D vision models, PCPT leverages less than 0.05 % of trainable parameters, while keeping the pre-trained parameters of the Transformer backbone unchanged. To evaluate the effectiveness of PCPT, extensive experiments were conducted on four discriminative datasets (ModelNet40, few-shot ModelNet40, ScanObjectNN, ShapeNetPart) and four generation datasets (PCN, MVP, ShapeNet55, and ShapeNet34/Unseen21). The results demonstrate that the task-specific prompts utilized in PCPT enable the Transformer model to adapt effectively to the target domains, yielding results comparable to those obtained through other full fine-tuning methods. This highlights the versatility of PCPT across various domains and tasks. Our code is available at https://***/Fayeben/PCPT. IEEE

关键词： Three dimensional displays

来源：评论

学校读者我要写书评

暂无评论

Numerical modelling of gas flow in a random packed bed with a helical flow deflector 26

Numerical modelling of gas flow in a random packed bed with ...

引用

26th Fluid Mechanics Conference, FMC 2024

作者： Marek, Maciej Faculty of Mechanical Engineering and Computer Science Czestochowa University of Technology Czestochowa Poland

Random packed beds are often employed in chemical reactors as a means to increase the contact surface between reactants or a catalyst. The present work proposes a helical flow deflector placed within the bed and numerical simulations are performed to analyse the influence of the deflector on the tortuosity of the flow paths in the reactor and the pressure drop. The geometry of the packed bed is obtained in a separate numerical simulation with sequential deposition of ring-like equilateral particles. The flow equations are solved on a structured grid and the immersed boundary method accounts for the complex geometry of the random packing and the cylindrical walls of the reactor. The increase of the tortuosity leads to increase of the pressure drop along the reactor and the relation between them is studied as a function of the rise per revolution of the helical surface. © Published under licence by IOP Publishing Ltd.

关键词： Pressure drop

来源：评论

学校读者我要写书评

暂无评论

Towards end-to-end structure determination from x-ray diffraction data using deep learning

引用

npj Computational Materials 2024年第1期10卷 1048-1059页

作者： Gabe Guo Judah Goldfeder Ling Lan Aniv Ray Albert Hanming Yang Boyuan Chen Simon J.L.Billinge Hod Lipson Department of Computer Science Columbia UniversityNew YorkNY10027USA Department of Applied Physics and Applied Mathematics Columbia UniversityNew YorkNY10027USA Department of Mechanical Engineering and Materials Science Duke UniversityDurhamNC27708USA Department of Mechanical Engineering Columbia UniversityNew YorkNY10027USA

Powder crystallography is the experimental science of determining the structure of molecules provided in crystalline-powder form,by analyzing their x-ray diffraction(XRD)*** many materials are readily available as crystalline powder,powder crystallography is of growing usefulness to many ***,powder crystallography does not have an analytically known solution,and therefore the structural inference typically involves a laborious process of iterative design,structural refinement,and domain knowledge of skilled experts.A key obstacle to fully automating the inference process computationally has been formulating the problem in an end-to-end quantitative form that is suitable for machine learning,while capturing the ambiguities around molecule orientation,symmetries,and reconstruction *** we present an ML approach for structure determination from powder diffraction *** works by estimating the electron density in a unit cell using a variational coordinate-based deep neural *** demonstrate the approach on computed powder x-ray diffraction(PXRD),along with partial chemical composition information,as *** evaluated on theoretically simulated data for the cubic and trigonal crystal systems,the system achieves up to 93.4%average similarity(as measured by structural similarity index)with the ground truth on unseen materials,both with known and partially-known chemical composition information,showing great promise for successful structure solution even from degraded and incomplete input *** approach does not presuppose a crystalline structure and the approach are readily extended to other situations such as nanomaterials and textured samples,paving the way to reconstruction of yet unresolved nanostructures.

关键词： structure powder crystalline

来源：评论

学校读者我要写书评

暂无评论

Progressive Growth for Point Cloud Completion by Surface-Projection Optimization

IEEE Transactions on Intelligent Vehicles

引用

IEEE Transactions on Intelligent Vehicles 2024年第5期9卷 1-15页

作者： Fei, Ben Zhang, Rui Yang, Weidong Li, Zhijun Chen, Wen-Ming School of Computer Science Fudan University Shanghai China School of Mechanical Engineering Tongji University Shanghai China Academy for Engineering and Technology Fudan University Shanghai China

Point cloud completion concentrates on completing geometric and topological shapes from incomplete 3D shapes. Nevertheless, the unordered nature of point clouds will hamper the generation of high-quality point clouds without predicting structured and topological information of the complete shapes and introducing noisy points. To effectively address the challenges posed by missing topology and noisy points, we introduce SPOFormer, a novel topology-aware model that utilizes surface-projection optimization in a progressive growth manner. SPOFormer consists of three distinct steps for completing the missing topology: (1) Missing Keypoints Prediction. A topology-aware transformer auto-encoder is integrated for missing keypoint prediction. (2) Skeleton Generation. The skeleton generation module produces a new type of representation named skeletons with the aid of keypoints predicted by topology-aware transformer auto-encoder and the partial input. (3) Progressively Growth. We design a progressive growth module to predict final output under Multi-scale Supervision and Surface-projection Optimization. Surface-projection Optimization is firstly devised for point cloud completion, aiming to enforce the generated points to align with the underlying object surface. Experimentally, SPOFormer model achieves an impressive Chamfer Distance-$\ell _{1}$ (CD) score of 8.11 on PCN dataset. Furthermore, it attains average CD-$\ell _{2}$ scores of 1.13, 1.14, and 1.70 on ShapeNet-55, ShapeNet-34, and ShapeNet-Unseen21 datasets, respectively. Additionally, the model achieves a Maximum Mean Discrepancy (MMD) of 0.523 on the real-world KITTI dataset. These outstanding qualitative and quantitative performances surpass previous approaches by a significant margin, firmly establishing new state-of-the-art performance across various benchmark datasets. Our code is available at https://***/kiddoray/SPOFormer IEEE

关键词： Topology

来源：评论

学校读者我要写书评

暂无评论

A systematic literature review of visual feature learning: deep learning techniques, applications, challenges and future directions

引用

Multimedia Tools and Applications 2024年 1-58页

作者： Abdullahi, Mohammed Oyelade, Olaide Nathaniel Kana, Armand Florentin Donfack Bagiwa, Mustapha Aminu Abdullahi, Fatimah Binta Junaidu, Sahalu Balarabe Iliyasu, Ibrahim Ore-ofe, Ajayi Chiroma, Haruna Department of Computer Science Ahmadu Bello University Zaria Zaria Nigeria Department of Mechanical Engineering Ahmadu Bello University Zaria Nigeria Department of Computer Engineering Ahmadu Bello University Zaria Nigeria College of Computer Science and Engineering University of Hafr Al Batin Hafar Al-Batin31991 Saudi Arabia School of Electronics Electrical Engineering and Computer Science Queen’s University Belfast Belfast United Kingdom

Visual Feature Learning (VFL) is a critical area of research in computer vision that involves the automatic extraction of features and patterns from images and videos. The applications of VFL are vast, including object detection and recognition, facial recognition, scene understanding, medical image analysis, and autonomous vehicles. In this paper, we propose to conduct extensive systematic literature review (SLR) on VFL based on deep learning algorithms. The paper conducted an SLR covering deep learning algorithms such as Convolutional Neural Networks (CNNs), Autoencoders, and Generative Adversarial Networks (GANs) including their variants. The review highlights the importance of VFL in computer vision and the limitations of traditional feature extraction techniques. Furthermore, it provides an in-depth analysis of the strengths and weaknesses of various deep learning algorithms for solving problems in VFL. The discussion of the applications of VFL provides an insight into the impact of VFL on various industries and domains. The review also analyzed the challenges faced by VFL, such as data scarcity and quality, overfitting, generalization, interpretability, and explainability. The discussion of future directions for VFL includes hybrid techniques, unsupervised feature learning, continual learning, attention-based models, and explainable AI. These techniques aim to address the challenges faced by VFL and improve the performance of the models. The systematic literature review concludes that VFL is a rapidly evolving field with the potential to transform many industries and domains. The review highlights the need for further research in VFL and emphasizes the importance of responsible use of VFL models in various applications. The review provides valuable insights for researchers and practitioners in the field of computer vision, who can use these insights to enhance their work and ensure the responsible use of VFL models. © The Author(s), under exclusive licence to

关键词： computer vision

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：