检索结果-内蒙古大学图书馆

Additive Manufacturing of Ti6Al4V Alloy Through Laser Metal Deposition (LMD): An Elebotative Approach of Process Modelling Via Computational Fluid Dynamics

学校读者我要写书评

暂无评论

SSRN

SSRN 2023年

作者： Rehman, Asif Ur Azher, Kashif Rehman, Aqsa Shah, Mussadiq Mahmood, Muhammed Arif Tüfekci, Celal Sami Salamci, Metin U. ERMAKSAN Bursa Turkey Department of Mechanical Engineering Faculty of Engineering Gazi University Ankara Turkey Additive Manufacturing Technologies Application and Research Center-EKTAM Ankara Turkey Department of Computer Science NFC-IET Punjab Multan60000 Pakistan Advanced Manufacturing Technologies Center of Excellence-URTEMM Ankara Turkey Intelligent Systems Center Missouri University of Science and Technology RollaMO65409 United States

Additive manufacturing (AM) has revolutionized many industries. Laser Melting Deposition (LMD) offers remarkable utility in size and is an effective method in terms of energy efficiency. The interaction of powder with laser before coming in contact with the substrate helps to attain fair melting and deposition. However, there is a problem associated with non-uniformity and accumulation of metallic melt on the edges of the manufactured parts which needs to be understood. Herein, LMD process of Ti6Al4V alloy is presented. The forces defining the melt pool flow have been elaborated. The phenomenon of recoil pressure and Marangoni actions along the single tracks of the melt pool is defined via the computational fluid dynamics (CFD) analysis. The behaviour of the melt pool in conduction mode and to check whether it will form a keyhole region or not. The main factor which influences the melt pool dynamics are found to be surface tension and density differences. Two modes of heat transfer conduction and depression are discussed with the Marangoni actions. Conduction mode is mainly governed by Marangoni actions while the depression mode is by recoil pressure. The stream traces and velocity vectors are observed for single tracks LMD process. A similar Marangoni actions experimental investigation is compared with the phenomenon observed under CFD analysis. The Discrete Element Modelling (DEM) adds value to the study as variable particle diameters are considered which are near to the actual powder size distribution. Moreover, considering the conduction mode flow dynamics to avoid the depression mode and generation of keyhole region in the melt pool will lead to porosity. © 2023, The Authors. All rights reserved.

关键词： Computational fluid dynamics

Two-Stream Regression Network for Dental Implant Position Prediction

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Yang, Xinquan Li, Xuguang Li, Xuechen Chen, Wenting Shen, Linlin Li, Xin Deng, Yongqiang College of Computer Science and Software Engineering Shenzhen University Shenzhen China AI Research Center for Medical Image Analysis and Diagnosis Shenzhen University Shenzhen China National Engineering Laboratory for Big Data System Computing Technology Shenzhen University China Department of Stomatology Shenzhen University General Hospital Shenzhen China Department of Electrical Engineering City University of Hong Kong Hong Kong

In implant prosthesis treatment, the design of surgical guide requires lots of manual labors and is prone to subjective variations. When deep learning based methods has started to be applied to address this problem, the space between teeth are various and some of them might present similar texture characteristic with the actual implant region. Both problems make a big challenge for the implant position prediction. In this paper, we develop a two-stream implant position regression framework (TSIPR), which consists of an implant region detector (IRD) and a multi-scale patch embedding regression network (MSPENet), to address this issue. For the training of IRD, we extend the original annotation to provide additional supervisory information, which contains much more rich characteristic and do not introduce extra labeling costs. A multi-scale patch embedding module is designed for the MSPENet to adaptively extract features from the images with various tooth spacing. The global-local feature interaction block is designed to build the encoder of MSPENet, which combines the transformer and convolution for enriched feature representation. During inference, the RoI mask extracted from the IRD is used to refine the prediction results of the MSPENet. Extensive experiments on a dental implant dataset through five-fold cross-validation demonstrated that the proposed TSIPR achieves superior performance than existing methods. Copyright © 2023, The Authors. All rights reserved.

关键词： Embeddings

Learning Hierarchical Graph Representation for Image Manipulation Detection

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Pan, Wenyan Zhou, Zhili Ling, Miaogen Geng, Xin Wu, Q.M. Jonathan School of Computer and Software & Engineering Research Center of Digital Forensics Ministry of Education Nanjing University of Information Science and Technology Nanjing China Provincial Key Laboratory for Computer Information Processing Technology Soochow University Suzhou China School of Computer Science and Engineering Southeast University Nanjing China Department of Electrical and Computer Engineering University of Windsor WindsorON Canada

The objective of image manipulation detection is to identify and locate the manipulated regions in the images. Recent approaches mostly adopt the sophisticated Convolutional Neural Networks (CNNs) to capture the tampering artifacts left in the images to locate the manipulated regions. However, these approaches ignore the feature correlations, i.e., feature inconsistencies, between manipulated regions and non-manipulated regions, leading to inferior detection performance. To address this issue, we propose a hierarchical Graph Convolutional Network (HGCN-Net), which consists of two parallel branches: the backbone network branch and the hierarchical graph representation learning (HGRL) branch for image manipulation detection. Specifically, the feature maps of a given image are extracted by the backbone network branch, and then the feature correlations within the feature maps are modeled as a set of fully-connected graphs for learning the hierarchical graph representation by the HGRL branch. The learned hierarchical graph representation can sufficiently capture the feature correlations across different scales, and thus it provides high discriminability for distinguishing manipulated and non-manipulated regions. Extensive experiments on four public datasets demonstrate that the proposed HGCN-Net not only provides promising detection accuracy, but also achieves strong robustness under a variety of common image attacks in the task of image manipulation detection, compared to the state-of-the-arts. Copyright © 2022, The Authors. All rights reserved.

关键词： Convolutional neural networks

FreezePipe: An Efficient Dynamic Pipeline Parallel Approach Based on Freezing Mechanism for Distributed DNN Training

学校读者我要写书评

暂无评论

FreezePipe: An Efficient Dynamic Pipeline Parallel Approach ...

International Conference on computer Supported Cooperative Work in Design

作者： Caishan Weng Zhiyang Shu Zhengjia Xu Jinghui Zhang Junzhou Luo Fang Dong Peng Wang Zhengang Wang School of Computer Science and Engineering Southeast University Nanjing P.R. China Boston College Nanjing P.R. China Data Science Center of Excellence Deloitte Consulting(Shanghai) Co. Ltd China Jiangsu HHigh-Tech Software Technology Co.Ltd

Deep Neural Network (DNN) training on a large scale is extremely time-consuming and computationally intensive, which is accelerated by distributed training. In recent years, pipeline parallelism has been developed, which enables partitioning the model across several devices, e.g. GPU, and training efficiency is improved by dividing data batches into micro-batches, with each of them processed by a different stage of the model. Currently, parallel training assumes pipeline placement and partitioning are static, with parameters updating each iteration, without accounting for freezing. This results in computational resources not being fully utilized. In this paper, we propose FreezePipe, a novel method for optimizing deep learning training that combines the freezing mechanism with pipeline parallel training. In FreezePipe, a lightweight method for determining the freezing strategy based on gradient changes is employed. Considering that resources need to be released based on the frozen layer, a lightweight model partitioning algorithm was designed to determine the optimal strategy for pipeline partitioning. Experimental results show that FreezePipe can reduce the training time by 64.5% compared to Torchgpipe on CIFAR-10 dataset without compromising any model performance.

关键词：

SAM-dPCR: Real-Time and High-throughput Absolute Quantification of Biological Samples Using Zero-Shot Segment Anything Model

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Wei, Yuanyuan Luo, Shanhang Xu, Changran Fu, Yingqi Zhang, Yi Qu, Fuyang Cheng, Guangyao Ho, Yi-Ping Ho, Ho-Pui Yuan, Wu Department of Biomedical Engineering The Chinese University of Hong Kong Shatin999077 Hong Kong Department of Biomedical Engineering National University of Singapore 119077 Singapore Department of Computer Science and Engineering The Chinese University of Hong Kong Shatin999077 Hong Kong Department of Electronic Engineering The Chinese University of Hong Kong Shatin999077 Hong Kong Centre for Biomaterials The Chinese University of Hong Kong 999077 Hong Kong Hong Kong Branch of CAS Center for Excellence in Animal Evolution and Genetics 999077 Hong Kong State Key Laboratory of Marine Pollution City University of Hong Kong 999077 Hong Kong

Digital PCR (dPCR) has revolutionized nucleic acid diagnostics by enabling absolute quantification of rare mutations and target sequences. However, current detection methodologies face challenges, as flow cytometers are costly and complex, while fluorescence imaging methods, relying on software or manual counting, are time-consuming and prone to errors. To address these limitations, we present SAM-dPCR, a novel self-supervised learning-based pipeline that enables real-time and high-throughput absolute quantification of biological samples. Leveraging the zero-shot SAM model, SAM-dPCR efficiently analyzes diverse microreactors with over 97.7% accuracy within a rapid processing time of 3.16 seconds. By utilizing commonly available lab fluorescence microscopes, SAM-dPCR facilitates the quantification of sample concentrations ranging from 0.74 to 17.49× 103 copies µL-1 for target nucleic acid templates. The accuracy of SAM-dPCR is validated by the strong linear relationship (R2 = 0.9939) observed between known and inferred sample concentrations. Additionally, SAM-dPCR demonstrates versatility through comprehensive verification using various samples and reactor morphologies. This accessible, cost-effective tool transcends the limitations of traditional detection methods or fully supervised AI models, marking the first application of SAM in nucleic acid detection or molecular diagnostics. By eliminating the need for annotated training data, SAM-dPCR holds great application potential for nucleic acid quantification in resource-limited settings. Copyright © 2024, The Authors. All rights reserved.

关键词： Fluorescence imaging

HierarchyNet: Learning to Summarize Source Code with Heterogeneous Representations

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Nguyen, Minh Huynh Bui, Nghi D.Q. Son, Hy Truong Thanh, Long Tran Nguyen, Tien N. FPT Software AI Center Department of Computer Science Fulbright University Viet Nam Department of Computer Science University of Chicago IL United States Department of Computer Science University of Warwick United Kingdom School of Engineering and Computer Science The University of Texas Dallas United States

We propose a novel method for code summarization utilizing Heterogeneous Code Representations (HCRs) and our specially designed HierarchyNet. HCRs effectively capture essential code features at lexical, syntactic, and semantic levels by abstracting coarse-grained code elements and incorporating fine-grained program elements in a hierarchical structure. Our HierarchyNet method processes each layer of the HCR separately through a unique combination of the Heterogeneous Graph Transformer, a Tree-based CNN, and a Transformer Encoder. This approach preserves dependencies between code elements and captures relations through a novel Hierarchical-Aware Cross Attention layer. Our method surpasses current state-of-the-art techniques, such as PA-Former, CAST, and NeuralCodeSum. Copyright © 2022, The Authors. All rights reserved.

关键词： Semantics

Using artificial intelligence to predict the next deceptive movement based on video sequence analysis: A case study on a professional cricket player's movements

学校读者我要写书评

暂无评论

Journal of Engineering Research (Kuwait)

Journal of engineering Research (Kuwait) 2025年

作者： Mutawa, A.M. Rajesh Kumar, Korupalli V. K, Hemachandran Murugappan, M. Computer Engineering Department College of Engineering and Petroleum Kuwait University Safat 5969 13060 Kuwait CoE Sports Analytics and Management and AI Research Centre School of Business Woxsen University Hyderabad India Intelligent Signal Processing (ISP) Research Lab Department of Electronics and Communication Engineering Kuwait College of Science and Technology Block 4 Doha 13133 Kuwait Department of Electronics and Communication Engineering School of Engineering Vels Institute of Sciences Technology and Advanced Studies Tamilnadu Chennai India Center of Excellence for Unmanned Aerial Systems (CoEUAS) Universiti Malaysia Perlis Perlis 02600 Malaysia

This research develops an artificial intelligence-based model to predict the next deceptive movement of athletes by analyzing video sequences of previous movements. Utilizing advanced deep neural network models, we analyze deceptive movements to forecast the next move, with a practical application on the deceptive movements of a professional cricket player. The model employs machine learning techniques such as Random Forest (RF), Decision Trees (DT), and K-Nearest Neighbor (KNN) to enhance prediction accuracy. Achieving up to 70 % accuracy, this model rivals human capability, as even highly skilled players can easily fall for deceptive actions. The ability to predict deceptive movements sets humans apart from many intelligent creatures, allowing athletes to avoid predictable actions and gain an edge in various sports. This study applies this concept to cricket, leveraging video data to improve training methods. The results highlight the potential of artificial intelligence in revolutionizing training and performance optimization in sports. © 2025 The Authors

关键词： Artificial intelligence Cricket Deceptive movements Machine learning Performance optimization

Fundus Image Translation with Scale-Aware Registration and Gradient-Guided GAN

学校读者我要写书评

暂无评论

SSRN

SSRN 2024年

作者： Kim, JuChan Le, Duc-Tai Bum, Junghyun Son, Chang-Hwan Lee, Eun Jung Han, Jong Chul Choo, Hyunseung Department of Artificial Intelligence Systems Engineering Sungkyunkwan University Korea Republic of College of Computing and Informatics Sungkyunkwan University Korea Republic of Sungkyun AI Research Institute Sungkyunkwan University Korea Republic of Software Science and Engineering Major Kunsan National University Korea Republic of Department of Ophthalmology Samsung Medical Center Korea Republic of School of Medicine Sungkyunkwan University Korea Republic of Department of Medical Device Management and Research SAIHST Sungkyunkwan University Korea Republic of Department of Electrical and Computer Engineering Sungkyunkwan University Korea Republic of

Two major fundus imaging types, conventional fundus (CF) imaging and ultrawidefield fundus (UF) imaging, are widely used in ophthalmology. In a typical clinical setting, both imaging methods are used simultaneously to diagnose diseases depending on the case. This study introduced an enhanced method of modality transformation to boost the diagnostic support capabilities of deep-learning models in ophthalmology. The method proposed in this study to translate UF images (UFIs) to CF images (CFIs) can replace the common clinical approach of both imaging methods, thereby reducing financial and temporal burdens on patients. This study learns and transforms the relationship between these modalities using UFI–CFI image pairs obtained from the same patient on the same day to minimize distorted information and accurately convert between different modalities. Our proposed model uses scale-aware registration and gradient-guided translation methods to align the UFI–CFI pairs, The generated CFIs not only enhance the quality of images and better represent the retinal area compared to existing methods but also effectively preserve the disease-related details in UFIs, aiding in the diagnosis. In addition, we optimize the image transformation performance, resulting in an 22% improvement in image quality assessment scales compared with previous methods. Furthermore, we acquired the observation that our results are nearly impossible to discern any differences between the real CFIs and the generated CFIs in the view of ophthalmologists. © 2024, The Authors. All rights reserved.

关键词： Ophthalmology

Noncontact Doppler Radar-based Heart Rate Detection on the SVD and ANC

学校读者我要写书评

暂无评论

Noncontact Doppler Radar-based Heart Rate Detection on the S...

2022 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2022

作者： Shi, Qiuxia Hu, Bin Tian, Fuze Zhao, Qinglin Gansu Provincial Key Laboratory of Wearable Computing School of Information Science and Engineering Lanzhou University China School of Computer and Communication Lanzhou University of Technology China Key Laboratory of Special Functional Materials and Structural Design Lanzhou University Ministry of Education Lanzhou China CAS Center for Excellence in Brain Science and Intelligence Technology Shanghai Institutes for Biological Sciences Chinese Academy of Sciences China Joint Research Center for Cognitive Neurosensor Technology of Lanzhou University Institute of Semiconductors Chinese Academy of Sciences China Engineering Research Center of Open Source Software and Real-Time System Lanzhou University Ministry of Education China

ISBN: (纸本)9781665468190

In the Doppler biological radar-based applications of noncontact measurement of vital signs, effectively extracting heartbeat information from weak thoracic mechanical motion is an important problem to be solved. This study is aimed at extracting heartbeat signal via the technology combined with Short Time Fourier Transform (STFT), Singular Value Decomposition (SVD) and Adaptive Noise Canceller (ANC) from radar recording. The simulated data and the data collected by Doppler radar biosensor realized in laboratory are employed to validate the proposed method. The results show that the proposed method has the ability of detection for the heart rate and heart rate variability indexes in rest state, it has certain advantages in time-consuming and detection accuracy. Therefore, the current method provides another way to process vital sign signals recorded by Doppler radar. © 2022 IEEE.

关键词： Singular value decomposition