检索结果-内蒙古大学图书馆

22nd IEEE International Symposium on Parallel and Distributed Processing with Applications, ISPA 2024

作者： Qin, Jiawei Wang, Yaobin Li, Ling Yang, Shuang Zhang, Xiaorong Southwest University of Science and Technology School of Computer Science and Technology Mianyang621010 China

ISBN: (纸本)9798331509712

Advanced Encryption Standard (AES), as one of the most popular encryption algorithms, has been widely studied on single GPU and CPU. However, the research on multi-GPU platforms is not deep enough, and with the rapid increase in data size, it is difficult for single-GPU platforms to meet the demand for high-performance computing. To solve this problem, this paper presents a novel AES parallelization study "P-AES"on MGPUSim, including its parallel execution mechanism as well as architectural design. In addition, we modify the instruction set of MGPUSim along with thread optimization to save memory overhead, increase data access speed, and improve system performance by loading S-boxes and key extension arrays from global memory to shared memory. The experimental results show that: (1) P-AES performs well on MGPUSim, achieving an average speedup of 1.5×-1.7× compared to the AES implementation using CUDA kernels, and a speedup of 2.75× compared to M-AES (the unoptimized AES on MGPUSim). (2) By conducting encryption experiments on plaintext data of varying sizes, we found that P-AES exhibits high stability and throughput. Compared to the modern sliced AES, the P-AES algorithm achieves a throughput of up to 812 Gbps, which is 1.3 times that of their implementation and 2.74 times that of M-AES. © 2024 IEEE.

关键词： computer graphics equipment

来源：评论

学校读者我要写书评

暂无评论

Accelerating Sparse Matrix-Matrix Multiplication by Adaptive Batching Strategy on MGPUSim 17

Accelerating Sparse Matrix-Matrix Multiplication by Adaptive...

引用

17th IEEE International Conference on Social Computing and Networking, SocialCom 2024

作者： Wang, Tianhai Wang, Yaobin Peng, Yutao Song, Yingchen Peng, Qian Tang, Pingping Southwest University of Science and Technology School of Computer Science and Technology Mianyang621010 China

ISBN: (纸本)9798331521097

Sparse Matrix-Matrix Multiplication (SpMM) is a widely used algorithm in Machine Learning, particularly in the increasingly popular Graph Neural Networks (GNNs). SpMM is an essential arithmetic operation in GNNs and has been parallelized on various platforms to accelerate GNN training. However, it has not been deeply studied on multi-GPU *** this work, we parallelize the SpMM algorithm on MGPUSim, including the parallel execution mechanism and architecture design. More importantly, we propose an adaptive batching strategy (ABS) to handle the irregular memory access of sparse matrices and allocate work-item resources efficiently. The strategy addresses issues such as scheduling overhead caused by too large work-groups, resource contention, and poor parallelism resulting from performance degradation due to too small work-groups. ABS improves the GPU's command processor scheduling speed, increases the efficiency of the incoming request rate, and optimizes work-group overhead. Finally, we conducted experiments on the i9-10900F CPU and NVIDIA RTX 3070 GPU using a set of matrices from SuiteSparse. Experimental results show that ABS achieves an average performance acceleration of 1.53× compared to the baseline and 1.40× compared to CUDA. Compared to the latest SpCaches approach used on FPGAs, our approach achieves a 1.12× relative performance improvement. © 2024 IEEE.

关键词： Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

Adaptive neural network control of a 2-DOF helicopter system considering input constraints and global prescribed performance

引用

science China(Information sciences) 2024年第7期67卷 224-239页

作者： Zhijia ZHAO Jiale WU Zhijie LIU We HE C.L.Philip CHEN School of Mechanical and Electrical Engineering Guangzhou University School of Intelligence Science and Technology and Key Laboratory of Intelligent Bionic Unmanned Systems of Ministry of Education University of Science and Technology Beijing School of Computer Science and Engineering South China University of Technology Pazhou Lab

In this study, an adaptive neural network(NN) control is proposed for nonlinear two-degree-offreedom(2-DOF) helicopter systems considering the input constraints and global prescribed ***, radial basis function NN(RBFNN) is employed to estimate the unknown dynamics of the helicopter system. Second, a smooth nonaffine function is exploited to approximate and address nonlinear constraint functions. Subsequently, a new prescribed function is proposed, and an original constrained error is transformed into an equivalent unconstrained error using the error transformation and barrier function transformation methods. The analysis of the established Lyapunov function proves that the controlled system is globally uniformly bounded. Finally, the simulation and experimental results on a constructed Quanser's test platform verify the rationality and feasibility of the proposed control.

关键词： adaptive NN control 2-DOF helicopter global prescribed performance input constraints

来源：评论

学校读者我要写书评

暂无评论

OCRBench: on the hidden mystery of OCR in large multimodal models

引用

science China(Information sciences) 2024年第12期67卷 23-35页

作者： Yuliang LIU Zhang LI Mingxin HUANG Biao YANG Wenwen YU Chunyuan LI Xu-Cheng YIN Cheng-Lin LIU Lianwen JIN Xiang BAI School of Artificial Intelligence and Automation Huazhong University of Science and Technology School of Electronic and Information Engineering South China University of Technology Microsoft Research School of Computer & Communication Engineering University of Science and Technology Beijing Institute of Automation Chinese Academy of Sciences School of Software Engineering Huazhong University of Science and Technology

Large models have recently played a dominant role in natural language processing and multimodal vision-language learning. However, their effectiveness in text-related visual tasks remains relatively unexplored. In this paper, we conducted a comprehensive evaluation of large multimodal models, such as GPT4V and Gemini, in various text-related visual tasks including text recognition, scene text-centric visual question answering(VQA), document-oriented VQA, key information extraction(KIE), and handwritten mathematical expression recognition(HMER). To facilitate the assessment of optical character recognition(OCR) capabilities in large multimodal models, we propose OCRBench, a comprehensive evaluation benchmark. OCRBench contains 29 datasets, making it the most comprehensive OCR evaluation benchmark available. Furthermore, our study reveals both the strengths and weaknesses of these models, particularly in handling multilingual text, handwritten text, non-semantic text, and mathematical expression *** importantly, the baseline results presented in this study could provide a foundational framework for the conception and assessment of innovative strategies targeted at enhancing zero-shot multimodal *** evaluation pipeline and benchmark are available at https://***/Yuliang-Liu/Multimodal OCR.

关键词： large multimodal model OCR text recognition scene text-centric VQA document-oriented VQA key information extraction handwritten mathematical expression recognition

来源：评论

学校读者我要写书评

暂无评论

Research on Quality Control of Electronic Medical Records Based on Feature Fusion Text Classification 24

Research on Quality Control of Electronic Medical Records Ba...

引用

5th International Conference on Artificial Intelligence and computer Engineering, ICAICE 2024

作者： Sun, Feng Huang, Xiaofang School of Computer Science and Technology Southwest University of Science and Technology Sichuan Mianyang China

ISBN: (纸本)9798400718007

In China's healthcare system, the quality of medical records, as an important medical document that records the complete course of a patient's illness, is related to medical safety and clinical research. Since the traditional manual review of medical records for quality control is inefficient and time-consuming, we design a text classification model based on feature fusion to detect defects in medical records. The model introduces the text classification output of Bidirectional Encoder Represen-tations from Transformers (BERT) which makes the model more effective in understanding the overall meaning of a sentence. We introduced Bidirectional Long and Short Term Memory Net-works and Text Convolutional Neural Networks in order to better capture the bi-directional semantic information in the text and to extract local features efficiently. Meanwhile, in order to extract the key features of the text and mitigate the degradation problem when the information is passed through the network, the model introduces the self-attention mechanism and residual connectivity. This feature fusion combining multiple techniques enables the model to better understand and classify the medical record defective text and improves the classification performance. The model was trained and tested with the Medical Record Defective Text dataset and the results showed an accuracy of 83.33%. © 2024 Copyright held by the owner/author(s). Publication rights licensed to ACM.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Implementation and Optimization of 8×8 Block Discrete Cosine Transform on MGPUSim 22

Implementation and Optimization of 8×8 Block Discrete Cosin...

引用

22nd IEEE International Symposium on Parallel and Distributed Processing with Applications, ISPA 2024

作者： Yang, Shuang Wang, Yaobin Li, Ling Qin, Jiawei Bi, Guotang Southwest University of Science and Technology School of Computer Science and Technology Mianyang621010 China

ISBN: (纸本)9798331509712

Discrete cosine transform for 8×8 block(DCT8×8) is widely used in image compression due to its high signal decorrelation rate. Current research for DCT is mainly focused on CPU and single GPU platforms, and the exploration of multi-GPU architectures is still insufficient. With the rapid increase of image data volume, the DCT8×8 algorithm on CPU and single GPU architectures can no longer meet the demand of efficient computation. Therefore, optimizing DCT8×8 algorithms on multi-GPU architectures becomes particularly important. To address this challenge, this paper explores the porting and optimization strategies of DCT algorithms on the new MGPUSim ***, we extend the instruction implementation of MG-PUSim to accomplish the porting of the original DCT8×8 algorithm (O-DCT), and then propose a kernel implementation strategy for optimizing the DCT8×8 algorithm by setting the size of the workgroup appropriately so that the kernel function can take advantage of the symmetry of the transformation matrix to reduce the redundant computation, which we refer to as the M-DCT. We conducted experiments on multiple grayscale images with different resolutions. The experimental results show that O-DCT performs well on MGPUSim, achieving an average 1.93× speedup ratio compared to the original DCT8×8 kernel (kernel1) in the CUDA SDK. Further experimental results show that M-DCT improves the execution efficiency on MGPUSim by 1.76× relative to O-DCT, and also achieves an average of 1.3× acceleration ratio compared to the optimized DCT8×8 kernel (kernel2) in the CUDA SDK, which improves the computational performance of the DCT8×8 algorithm on multi-GPU architectures. It provides new ideas and methods for future research on image compression on multi-GPU architectures. © 2024 IEEE.

关键词： computer graphics equipment

来源：评论

学校读者我要写书评

暂无评论

Collaborative Cloud-Edge Computing with Mixed Wireless and Wired Backhaul Links: Joint Task Offloading and Resource Allocation 19th

Collaborative Cloud-Edge Computing with Mixed Wireless and...

引用

19th EAI International Conference on Collaborative Computing: Networking, Applications and Worksharing, CollaborateCom 2023

作者： Zhang, Daqing Sun, Haifeng School of Computer Science and Technology Southwest University of Science and Technology Mianyang621010 China

ISBN: (纸本)9783031545207

Mobile Edge Computing (MEC) is a promising technology that provides computing services at the edge of wireless networks to reduce the latency and the energy consumption for Smart Mobile Devices (SMDs). Additionally, the Ultra-Dense Network (UDN) will play a key role in providing high transmission capacity for SMDs in 5G networks. In order to improve the edge cloud efficiency within limited communication and computing resources, this paper proposes a joint task offloading and resource allocation scheme collaborated between cloud computing and edge computing in the UDN. Since wireless backhaul is more economical than expensive wired backhaul deployments, we consider the mixed deployment of either wired or wireless backhaul between each Small Base Station (SBS) and the Macro Base Station (MBS) in UDN scenarios, then formulate an optimization problem to minimize the system-wide computation overhead, and apply the Linear Decreasing Weight Particle Swarm Optimization (LDWPSO) algorithm to solve the problem. Numerical experiments validate the effectiveness of our proposed scheme compared to other baseline schemes. © ICST Institute for computer sciences, Social Informatics and Telecommunications Engineering 2024.

关键词： Resource allocation

来源：评论

学校读者我要写书评

暂无评论

Revealing the Fairness Issues of Text-to-Image Generation Models Based on Metamorphic Testing 17

Revealing the Fairness Issues of Text-to-Image Generation Mo...

引用

17th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics, CISP-BMEI 2024

作者： Ruan, MengYao Pan, Ya Fan, Yong School of Computer Science and Technology Southwest University of Science and Technology Mian Yang China

ISBN: (纸本)9798331507398

The Text-to-Image Generation(T2I) Models acquires implicit social biases during the training process, which can easily cause social disputes and negative impacts in sensitive fields such as news broadcasting, educational illustrations, and so on. There are many factors contributing to this situation. Most of the existing studies only focus on the combed biases of gender and skin color. Therefore, this paper proposes a method based on metamorphosis test to reveal the fairness of T2I models in terms of gender, skin color, and region. The part of speech and influential positions are divided in the input text, and the prompt text is modified by using names with regional characteristics in combination with three mutation operators: entity replacement, entity attribute enhancement, and multilingual description. The large language model is used for optimization to expand the test cases for generating images. To alleviate the test oracle problem of the T2I models, three metamorphic relations are constructed for verification in terms of gender, skin color, and image-text consistency. This paper has conducted a large number of experiments on T2I models: Kolors, Wan Xiang, and Stable Diffusion. The experimental results show that after variation using the variation operator, the proportion of female entities increases by at least 11.4%, and the proportion of dark-skinned objects increases by at least 15.7%. © 2024 IEEE.

关键词： Economic and social effects

来源：评论

学校读者我要写书评

暂无评论

LA-TransUNet: Intracranial Hemorrhage CT Image Segmentation Network Based on Attention Mechanism 17

LA-TransUNet: Intracranial Hemorrhage CT Image Segmentation ...

引用

17th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics, CISP-BMEI 2024

作者： Xu, Hongli Wu, Jue School of Computer Science and Technology Southwest University of Science and Technology Mian Yang China

ISBN: (纸本)9798331507398

Intracranial hemorrhage(ICH) is a serious disease with high morbidity, recurrence, and disability rates, and computed tomography (CT) is considered an important standard for diagnosis. Since CT images of ICH have problems such as small bleeding areas, blurred boundaries, irregular shapes, and high noise, segmentation using separate networks such as U-Net or TransUNet often fails to obtain accurate results. To solve this problem, we propose an ICH CT image segmentation model (LA-TransUNet) that combines the attention mechanism (SEpro), large kernel convolution selection module (LSK-Block), Transformer, and U-Net. The network mainly contains three core components: encoder, decoder, and skip connection. The encoder introduced SEpro to enhance feature extraction capabilities and reduce the impact of noise interference. SEpro is introduced in the skip connection to improve the model's ability to capture edge detail information. The decoder introduced LSK-Block to focus on important features to enhance the model's perception of bleeding areas. Experimental results show that LA-TransUNet is superior to other mainstream segmentation networks regarding segmentation results and evaluation indicators. © 2024 IEEE.

关键词： Image segmentation

来源：评论

学校读者我要写书评

暂无评论

Mobility-Aware Graph Reinforcement Learning for Service Migration in Mobile Edge Computing 17

Mobility-Aware Graph Reinforcement Learning for Service Migr...

引用

17th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics, CISP-BMEI 2024

作者： Liu, Shilong Sun, Haifeng School of Computer Science and Technology Southwest University of Science and Technology Mianyang621010 China

ISBN: (纸本)9798331507398

Mobile edge computing improves data processing efficiency and reduces latency by deploying computing and storage resources at the network edge, making it suitable for real-time applications. In vehicular networks, due to the high mobility of vehicles and the limited coverage of edge servers, ensuring Quality of Service (QoS) and preventing service inter-ruptions are critical challenges. Service migration and resource reallocation are necessary strategies to maintain QoS. One major challenge in MEC scenarios is how to deliver stable services in the face of high vehicle mobility. To address this, this paper proposes a Mobility-Aware Graph Reinforcement Learning (MA-G RL) framework, designed specifically for service migration in vehicular network systems. The MA-GRL framework consists of two components: the first is a vehicle position prediction module based on a sequence-to-sequence (seq2seq) model, and the second is a graph attention-based reinforcement learning (GRL) module for service migration decision-making and resource allocation. MA-GRL models the vehicular network as a graph and the service migration process as a Markov Decision Process (MDP), utilizing a graph attention network to handle the dynamic observation space and leveraging attention mechanisms to make decisions in a constantly changing action space. Simulation results show that MA-GRL outperforms traditional methods in reducing communication latency. Additionally, the trained model demonstrates adaptability and stability across different network topologies, highlighting its robustness in various environments. © 2024 IEEE.

关键词： Markov processes

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：