检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Tang, Ze Ge, Jidong Liu, Shangqing Zhu, Tingwei Xu, Tongtong Huang, Liguo Luo, Bin National Key Laboratory for Novel Software Technology Nanjing University China Nanyang Technological University Singapore Huawei Software Engineering Application Technology China Department of Computer Science Southern Methodist University United States

Large Language Models (LLMs) have demonstrated remarkable performance in code completion. However, due to the lack of domain-specific knowledge, they may not be optimal in completing code that requires intensive domain knowledge for example completing the library names. Although there are several works that have confirmed the effectiveness of fine-tuning techniques to adapt language models for code completion in specific domains. They are limited by the need for constant fine-tuning of the model when the project is in constant iteration. To address this limitation, in this paper, we propose kNM-LM, a retrieval-augmented language model (R-LM), that integrates domain knowledge into language models without fine-tuning. Different from previous techniques, our approach is able to automatically adapt to different language models and domains. Specifically, it utilizes the in-domain code to build the retrieval-based database decoupled from LM, and then combines it with LM through Bayesian inference to complete the code. The extensive experiments on the completion of intra-project and intra-scenario have confirmed that kNM-LM brings about appreciable enhancements when compared to CodeGPT and UnixCoder. A deep analysis of our tool including the responding speed, storage usage, specific type code completion, and API invocation completion has confirmed that kNM-LM provides satisfactory performance, which renders it highly appropriate for domain adaptive code completion. Furthermore, our approach operates without the requirement for direct access to the language model's parameters. As a result, it can seamlessly integrate with black-box code completion models, making it easy to integrate our approach as a plugin to further enhance the performance of these models. Copyright © 2023, The Authors. All rights reserved.

关键词： Domain Knowledge

来源：评论

学校读者我要写书评

暂无评论

ALCencryption:A Secure and Efficient Algorithm for Medical Image Encryption

引用

computer Modeling in Engineering & Sciences 2020年第12期125卷 1083-1100页

作者： Jiao Ge College of Computer Science and Technology Hengyang Normal UniversityHengyang421002China Hunan Provincial Key Laboratory of Intelligent Information Processing and Application Hengyang421002China

With the rapid development of medical informatization and the popularization of digital imaging equipment,DICOM images contain the personal privacy of patients,and there are security risks in the process of storage and transmission,so it needs to be *** order to solve the security problem of medical images on mobile devices,a safe and efficient medical image encryption algorithm called ALCencryption is *** algorithm first analyzes the medical image and distinguishes the color image from the gray *** gray images,the improved Arnold map is used to scramble them according to the optimal number of iterations,and then the diffusion is realized by the Logistic and Chebyshev map cross-diffusion *** color image is encrypted by cross-diffusion algorithm of double chaotic *** and efficiency analysis show that the ALCencryption algorithm has the characteristics of small neighboring pixels,large key space,strong key sensitivity,high safety and short encryption *** is suitable for medical image encryption of mobile devices with high real-time requirements.

关键词： Patient privacy DICOM medical image encryption scrambling degree cross-diffusion

来源：评论

学校读者我要写书评

暂无评论

Modeling Fine-Grained Relations in Dynamic Space-Time Graphs for Video-Based Facial Expression Recognition

引用

IEEE Transactions on Affective Computing 2025年

作者： Huang, Changqin Jiang, Fan Han, Zhongmei Huang, Xiaodi Wang, Shijin Zhu, Yanlai Jiang, Yunliang Hu, Bin Zhejiang Normal University Key Laboratory of Intelligent Education Technology and Application of Zhejiang Province Zhejiang Jinhua321004 China Charles Sturt University School of Computing Mathematics and Engineering AlburyNSW2640 Australia State Key Laboratory of Cognitive Intelligence Anhui Hefei230088 China Hangzhou Hikvision Digital Technology Co. Ltd Education Business Department Hangzhou310051 China Zhejiang Normal University School of Computer Science and Technology Jinhua321004 China Beijing Institute of Technology School of Medical Technology Beijing10081 China

Facial expressions in videos inherently mirror the dynamic nature of real-world facial events. Consequently, facial expression recognition (FER) should employ a dynamic graph-based representation to effectively capture the relational structure of facial expressions rather than relying on conventional grid or sequence methods. However, existing graph-based approaches have their limitations. Frame-level graph methods provide a coarse representation of the facial graph across time and space, while landmark-based graph methods need to introduce additional facial landmarks, resulting in a static graph structure. To address these challenges, we propose spatial-temporal relation-aware dynamic graph convolutional networks (ST-RDGCN). This fine-grained relation modeling approach enables the dynamic modeling of evolving facial expressions in videos through dynamic space-time graphs, eliminating the need for facial landmarks. ST-RDGCN encompasses three graph construction paradigms: dynamic independent space graph, dynamic joint space-time graph, and dynamic cross space-time graph. Furthermore, we propose a relation-aware space-time graph convolution (RSTG-Conv) operator to learn informative spatiotemporal correlations in dynamic space-time graphs. In extensive experimental evaluations, our ST-RDGCN demonstrates state-of-the-art performance on the five popular video-based FER datasets, achieving overall accuracy scores of 99.69%, 91.67%, 56.51%, 69.37%, and 49.03% on the CK+, Oulu-CASIA, AFEW, DFEW, and FERV39 k datasets, respectively. In particular, our ST-RDGCN outperforms the current best method by 3.6% in UAR on the most challenging FERV39 k dataset. Furthermore, our analysis reveals that the dynamic cross space-time graph scheme is the most effective among the three dynamic graph construction schemes. © 2010-2012 IEEE.

关键词： Graphic methods

来源：评论

学校读者我要写书评

暂无评论

Optimized Defogging Algorithm for Dark Channel Based on Multi-Scale Convolutional Neural Network

Optimized Defogging Algorithm for Dark Channel Based on Mult...

引用

Smart Grid and Artificial Intelligence (SGAI), International Conference on

作者： Zhaochuan Wei Rongtian Yang Somsack Inthasone Jundao Pan Guangxi Key Laboratory of Precision Navigation Technology and Application Guilin University of Electronic Technology Guilin China Computer Science Department Faculty of Natural Nayional University of Laos Vientiane Capital Lao PDR Chinese Academy of Sciences Institute of Space and Astronautical Information Innovation Beijing China

ISBN: (数字)9798331518806

ISBN: (纸本)9798331518813

To address the issue of significant discrepancies between the estimated transmission in traditional single-image dehazing algorithms and the actual scenario, a multi-scale convolutional neural network-based dehazing algorithm optimized by dark channel prior is proposed. The algorithm utilizes guided filtering, which preserves edge smoothness, to filter the feature transmission map extracted by the convolutional neural network, making the processed transmission closer to the actual transmission of the input hazy image. Subsequently, the dark channel prior theory is employed to obtain the global atmospheric light value of the hazy image, and finally, The atmospheric scattering model serves as a crucial tool when engaging in the operation of image dehazing, effectively enhancing the clarity of images affected by haze. The outcomes of the experiments indicate that this algorithm surpasses the latest approaches in both qualitative and quantitative evaluations. The algorithm put forward shows remarkable enhancements in peak signal - to - noise ratio and structural similarity, leading to significantlyenhanced dehazing performance.

关键词： Filtering Atmospheric modeling Noise Channel estimation Scattering Estimation Filtering algorithms Feature extraction Filtering theory Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Enhanced Fine-Tuning of Lightweight Domain-Specific Q&A Model Based on Large Language Models

arXiv

引用

arXiv 2024年

作者： Zhang, Shenglin Zhu, Pengtian Ma, Minghua Wang, Jiagang Sun, Yongqian Li, Dongwen Wang, Jingyu Guo, Qianying Hua, Xiaolei Zhu, Lin Pei, Dan Nankai University China Microsoft United States Tsinghua University China China Mobile Research Institute China Haihe Laboratory of Information Technology Application Innovation China Tianjin Key Laboratory of Software Experience and Human Computer Interaction China Beijing National Research Center for Information Science and Technology China

Large language models (LLMs) excel at general question-answering (Q&A) but often fall short in specialized domains due to a lack of domain-specific knowledge. Commercial companies face the dual challenges of privacy protection and resource constraints when involving LLMs for fine-tuning. This paper propose a novel framework, Self-Evolution, designed to address these issues by leveraging lightweight open-source LLMs through multiple iterative fine-tuning rounds. To enhance the efficiency of iterative fine-tuning, Self-Evolution employ a strategy that filters and reinforces the knowledge with higher value during the iterative process. We employed Self-Evolution on Qwen1.5-7B-Chat using 4,000 documents containing rich domain knowledge from China Mobile, achieving a performance score 174% higher on domain-specific question-answering evaluations than Qwen1.5-7B-Chat and even 22% higher than Qwen1.5-72B-Chat. Self-Evolution has been deployed in China Mobile's daily operation and maintenance for 117 days, and it improves the efficiency of locating alarms, fixing problems, and finding related reports, with an average efficiency improvement of over 18.6%. In addition, we release Self-Evolution framework code in https://***/Zero-Pointer/Self-Evolution. Copyright © 2024, The Authors. All rights reserved.

关键词： Data mining

来源：评论

学校读者我要写书评

暂无评论

Cognizant Resource Balancing in Montgomery Modular Multiplication Optimization for Dsp

SSRN

引用

SSRN 2024年

作者： Tao, Qiqi Li, Liying Zhou, Junlong Cao, Guitao Meng, Dan Shanghai Key Laboratory of Trustworthy Computing MoE Engineering Research Center of SW/HW Co-design Technology and Application East China Normal University Shanghai200062 China School of Computer Science and Engineering Nanjing University of Science and Technology Nanjing210094 China

The performance of modular multiplication directly influences the efficiency of a majority of encryption ***, there are numerous FPGA-based acceleration techniques targeting modular ***, many of these implementations require substantial hardware resources or suffer from resource ***, we present a novel FPGA-based implementation of Montgomery Modular Multiplication aimed at addressing these *** design employs a suitable radix bit width and word size based on the digital signal processing (DSP) bit width rather than the conventional binary powers of *** aim to instantiate more modular multipliers using limited resources while minimizing *** also introduce a novel DSP cascade structure, called parallel grouping cascade DSP, which reduces the number of clock cycles of internal *** balance the ratio of lookup table (LUT) and DSP usage, we also use multipliers implemented in the LUT to replace some *** results, implemented on Xilinx Virtex-7 field-programmable gate array (FPGA), demonstrate more than 27\% improvement in throughput on 1024-bit modular multiplication and more than 70\% improvement on 2048-bit compared to the best previous state-of-the-art references. © 2024, The Authors. All rights reserved.

关键词： Digital signal processing

来源：评论

学校读者我要写书评

暂无评论

Teaching resource recommendation for OBE based on kernel canonical correlation analysis 12

Teaching resource recommendation for OBE based on kernel can...

引用

12th International Conference on Information technology in Medicine and Education, ITME 2022

作者： Changqian, Wu Shaozi, Li Yaojin, Lin Xiehua, Yu MinNan Science and Technology Key Laboratory of Big Data and Artificial Intelligence University José Rizal University University of Fujian Province Quanzhou China Xiamen University School of Informatics Xiamen China Minnan Normal University School of Computer Science Lab of Data Science and Intelligence Application Zhangzhou China MinNan Science and Technology University Key Laboratory of Big Data and Artificial Intelligence University of Fujian Province Quanzhou China

ISBN: (纸本)9798350310153

Data mining is used to extract information from OBE teaching resources, then the teaching resources and user personality database are established to improve the effectiveness of resources recommendation on the online learning platform. The properties of teaching resources and user personality traits are analyzed using the kernel canonical correlation analysis algorithm, and teaching materials with a high correlation coefficient are chosen to recommend to users. The simulation results demonstrate that this proposed method has higher accuracy and more accurate recommended resources when compared to the widely used teaching resource recommendation techniques. © 2022 IEEE.

关键词： Data mining

来源：评论

学校读者我要写书评

暂无评论

Research on Two-stage Conveying Bar Counting and Dividing System Based on Vision

Research on Two-stage Conveying Bar Counting and Dividing Sy...

引用

第二十一届分布式计算及其应用国际学术研讨会

作者： Guojun Chen Yang Wu Jiangsu Key Laboratory of IoT Application Technology Wuxi Taihu University College of Computer Internet of Tings Engineering Wuxi Taihu University

A vision-based automatic bar counting system for two-stage conveying bars is *** system solves the counting problems of sticking and relative sliding of a large number of rebar stacks through image processing *** the improvement of the splitting end splitting machine,the design of two-step splitting and visual verification is adopted to achieve accurate splitting of stacked *** system obviously reduces the requirements for installation space,and can make more effective use of the space in the rebar *** system has fast running speed,high accuracy,small changes to the production line,and easy *** can not only meet the needs of production,but also greatly improve the efficiency of the existing counting *** is suitable for rebar workshops with limited installation *** experimental results show that the system can accurately count the bars and can accurately realize the steel *** bars with a diameter of 8-12mm,the counting accuracy is over 99.85%,and for bars with a diameter greater than 12mm,the counting accuracy is over 99.96%.

关键词： bar counting steel splitting pit splitting matching

来源：评论

学校读者我要写书评

暂无评论

ZuantuSet: A Collection of Historical Chinese Visualizations and Illustrations 25

ZuantuSet: A Collection of Historical Chinese Visualizations...

引用

2025 CHI Conference on Human Factors in Computing Systems, CHI 2025

作者： Mei, Xiyao Zhang, Yu Yang, Chaofan Shi, Rui Yuan, Xiaoru Peking University Beijing China University of Oxford Oxford United Kingdom National Key Laboratory of General Artificial Intelligence School of Intelligence Science and Technology Peking University China Department of Computer Science University of Oxford United Kingdom Center for Research on Ancient Chinese History Peking University China National Engineering Laboratory for Big Data Analysis and Application Peking University China PKU-WUHAN Institute for Artificial Intelligence China

ISBN: (纸本)9798400713941

Historical visualizations are a valuable resource for studying the history of visualization and inspecting the cultural context where they were created. When investigating historical visualizations, it is essential to consider contributions from different cultural frameworks to gain a comprehensive understanding. While there is extensive research on historical visualizations within the European cultural framework, this work shifts the focus to ancient China, a cultural context that remains underexplored by visualization researchers. To this aim, we propose a semi-automatic pipeline to collect, extract, and label historical Chinese visualizations. Through the pipeline, we curate ZuantuSet, a dataset with over 71K visualizations and 108K illustrations. We analyze distinctive design patterns of historical Chinese visualizations and their potential causes within the context of Chinese history and culture. We illustrate potential usage scenarios for this dataset, summarize the unique challenges and solutions associated with collecting historical Chinese visualizations, and outline future research directions. © 2025 Copyright held by the owner/author(s). Publication rights licensed to ACM.

关键词： data labeling dataset digital humanities historical visualization

来源：评论

学校读者我要写书评

暂无评论

Teaching Text-to-Image Models to Communicate in Dialog

arXiv

引用

arXiv 2023年

作者： Sun, Xiaowen Feng, Jiazhan Wang, Yuxuan Lai, Yuxuan Shen, Xingyu Zhao, Dongyan Wangxuan Institute of Computer Technology Peking University China National Key Laboratory of General Artificial Intelligence BIGAI China Engineering Research Center of Integration and Application of Digital Learning Technology Ministry of Education China Department of Computer Science The Open University of China China

A picture is worth a thousand words, thus, it is crucial for conversational agents to understand, perceive, and effectively respond with pictures. However, we find that directly employing conventional image generation techniques is inadequate for conversational agents to produce image responses effectively. In this paper, we focus on the innovative dialog-to-image generation task, where the model synthesizes a high-resolution image aligned with the given dialog context as a response. To tackle this problem, we design a tailored fine-tuning approach on the top of state-of-the-art text-to-image generation models to fully exploit the structural and semantic features in dialog context during image generation. Concretely, we linearize the dialog context with specific indicators to maintain the dialog structure, and employ in-domain data to alleviate the style mismatch between dialog-to-image and conventional image generation tasks. Empirical results on PhotoChat and MMDialog Corpus show that our approach brings consistent and remarkable improvement with 3 state-of-the-art pre-trained text-to-image generation backbones. Copyright © 2023, The Authors. All rights reserved.

关键词： Semantics

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：