检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Ma, Caoyuan Liu, Yu-Lun Wang, Zhixiang Liu, Wu Liu, Xinchen Wang, Zheng National Engineering Research Center for Multimedia Software Institute of Artificial Intelligence School of Computer Science Wuhan University China National Yang Ming Chiao Tung University Taiwan The University of Tokyo Japan National Institute of Informatics Japan School of Information Science and Technology University of Science and Technology of China China JD Explore Academy

We present HumanNeRF-SE, a simple yet effective method that synthesizes diverse novel pose images with simple input. Previous HumanNeRF works require a large number of optimizable parameters to fit the human images. Instead, we reload these approaches by combining explicit and implicit human representations to design both generalized rigid deformation and specific non-rigid deformation. Our key insight is that explicit shape can reduce the sampling points used to fit implicit representation, and frozen blending weights from SMPL constructing a generalized rigid deformation can effectively avoid overfitting and improve pose generalization performance. Our architecture involving both explicit and implicit representation is simple yet effective. Experiments demonstrate our model can synthesize images under arbitrary poses with few-shot input and increase the speed of synthesizing images by 15 times through a reduction in computational complexity without using any existing acceleration modules. Compared to the state-of-the-art HumanNeRF studies, HumanNeRF-SE achieves better performance with fewer learnable parameters and less training time. Copyright © 2023, The Authors. All rights reserved.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

Report-Guided Cross-Modal Representation Learning for Predicting EGFR Mutations by Whole Slide Image

Report-Guided Cross-Modal Representation Learning for Predic...

引用

IEEE International Conference on Bioinformatics and Biomedicine (BIBM)

作者： Qi Qiao Jun Shi Zhiguo Jiang Wei Wang Haibo Wu Yushan Zheng School of Computer Science and Information Engineering Hefei University of Technology Hefei China School of Software Hefei University of Technology Hefei China Image Processing Center School of Astronautics Beihang University Beijing China Department of Pathology the First Affiliated Hospital of USTC Division of Life Sciences and Medicine University of Science and Technology of China Hefei China Intelligent Pathology Institute Division of Life Sciences and Medicine University of Science and Technology of China Hefei China School of Engineering Medicine Beijing Advanced Innovation Center for Biomedical Engineering Beihang University Beijing China

ISBN: (数字)9798350386226

ISBN: (纸本)9798350386233

Traditional PCR/NGS-based multigene panel testing is time-consuming and costly. Predicting EGFR mutations directly from H&E stained whole slide images (WSIs) can alleviate these limitations. Furthermore, histopathological reports contain valuable textual information that correlates with tissue areas in WSIs. However, recent research mainly analyses EGFR mutation status only from a single modality, ignoring rich information contained in reports. In this paper, we propose a report-guided cross-modal representation learning method for predicting EGFR mutations by WSIs. Specifically, we reconstruct report-level embeddings through exploring intrinsic relationships between diagnostic words in histopathological reports and tissue areas in WSIs. Finally, reconstructed histopathological report embedding and aggregated WSI embedding are fused for final prediction. More importantly, molecular testing report is also introduced as prior supervision information at the training stage to guarantee semantic consistency of fused feature and molecular report embedding. We evaluate our method on the TCGA-EGFR public benchmark dataset and an in-house clinical dataset (USTC-EGFR). Experimental results demonstrate that our method outperforms existing approaches in EGFR mutation prediction, highlighting the benefits of cross-modal learning in enhancing feature representational ability. The code is available at https://***/HFUT-miaLab/RCRL.

关键词： Representation learning Training Codes Supervised learning Semantics Medical treatment Benchmark testing Bioinformatics Image reconstruction Biomedical imaging

来源：评论

学校读者我要写书评

暂无评论

The good, the bad, and why: unveiling emotions in generative AI 24

The good, the bad, and why: unveiling emotions in generative...

引用

Proceedings of the 41st International Conference on Machine Learning

作者： Cheng Li Jindong Wang Yixuan Zhang Kaijie Zhu Xinyi Wang Wenxin Hou Jianxun Lian Fang Luo Qiang Yang Xing Xie Microsoft Research Beijing China and CAS Institute of Software Beijing China Microsoft Research Beijing China Department of Computer Science William & Mary Williamsburg Virginia America School of Psychology Beijing Normal University Beijing China Department of Computer Science and Engineering Hong Kong University of Science and Technology Hong Kong China

Emotion significantly affects our daily behaviors and interactions. Although recent generative AI models, such as large language models, have shown impressive performance in various tasks, it remains unclear whether they truly comprehend emotions and why. This paper aims to address this gap by incorporating psychological theories to gain a holistic understanding of emotions in generative AI models. Specifically, we propose three approaches: 1) EmotionPrompt to enhance the performance of the AI model, 2) EmotionAttack to impair the performance of the AI model, and 3) EmotionDecode to explain the effects of emotional stimuli, both benign and malignant. Through extensive experiments involving language and multi-modal models on semantic understanding, logical reasoning, and generation tasks, we demonstrate that both textual and visual EmotionPrompt can boost the performance of AI models while EmotionAttack can hinder it. More importantly, EmotionDecode reveals that AI models can comprehend emotional stimuli similar to the dopamine mechanism in the human brain. Our work heralds a novel avenue for exploring psychology to enhance our understanding of generative AI models, thus boosting the research and development of human-AI collaboration and mitigating potential risks.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Fault Diagnosis Model of Switching Power Supply Based on Improved DBN

Fault Diagnosis Model of Switching Power Supply Based on Imp...

引用

International Conference on Intelligent System and Knowledge engineering, ISKE

作者： Lei Zhou Cheng Yu Hu Liu Xin Li Fenfen Li Haiyan Zhang Faculty of Computer and Software Engineering Research Center for logic and Intelligent Computation National-Local Joint Engineering Lab of System Credibility Automatic Verification Huaiyin Institute of Technology Huai’an Jiangsu China College of Computer Science and Technology Nanjing University of Technology Nanjing Jiangsu China Faculty of Chemical Engineering Huaiyin Institute of Technology Huai’an Jiangsu China

ISBN: (纸本)9781665405546

Switching power supply has wide application in various electrical equipment. The failure of high-voltage filter capacitors is the main cause of frequent faults, accounting for more than 60% of switching power supply failures. In order to realize the real-time detection of switching power supply failure without disassembling the power supply, a fault diagnosis model for improving the deep belief network (DTDBN) is proposed. The model uses the data acquisition device to isolate, filter, and collect data on the grid voltage of the switching power supply, judge the failure ratio of the filter capacitor according to the degree of voltage distortion of the power supply. Experimental results show that the proposed fault detection method achieves a 100% recognition rate for the filter capacitor normal state, 50% and 90% state recognition rate is 90%. The superiority of this method for the classification and early warning of power capacitor faults is proved.

关键词： Fault diagnosis Power supplies Fault detection Capacitors Switches Voltage Distortion

来源：评论

学校读者我要写书评

暂无评论

Research of Sampling Method For Cybercrime Black Industry Chain

Research of Sampling Method For Cybercrime Black Industry Ch...

引用

IEEE International Conference on Electronic Information and Communication technology (ICEICT)

作者： Ying’An Cui Xue Li Yande Wu Department of Computer Science and Engineering Xi’an University of Technology Xi’an China School of International Business Shaanxi Normal University Xi’an China Software R&D Department 2 China Unicom Western Innovation Research Institute Co. Ltd Xi’an China

ISBN: (数字)9798350384437

ISBN: (纸本)9798350384444

In modern organization theories, the social network established by cyber criminals is termed the crime cyber black industry chain. Under the current technical level, treating a crime cyber black industry chain as a whole dataset has to face severe challenges in many aspects, so the research of the sampling method specifically for it has been one of the current hot topics. Based on the structural feature of a cybercrime industry chain, the overall framework and the detailed algorithm of a multi-stage sampling method are proposed. Tested with real cybercrime datasets, our proposed multi-stage sampling method outperforms existing mainstream single-stage sampling methods at macro, meso, and micro scales, indicating it can solve practical problems with high efficiency and low cost.

关键词： Industries Costs Social networking (online) Organizations Sampling methods Information and communication technology Reliability

来源：评论

学校读者我要写书评

暂无评论

Intellicare: Improving Healthcare Analysis with Patient-Level Knowledge from Large Language Models

SSRN

引用

SSRN 2025年

作者： Yu, Zhihao Jin, Yujie Xu, Yongxin Zhang, Xiaoyun Wang, Yasha School of Computer Science Peking University Beijing100091 China National Research and Engineering Center of Software Engineering Beijing100091 China School and Hospital of Stomatology National Center for Stomatology National Clinical Research Center for Oral Diseases National Engineering Research Center of Oral Biomaterials and Digital Medical Devices Beijing100081 China Key Laboratory of High Confidence Software Technologies Ministry of Education Beijing100091 China Peking University Information Technology Institute Tianjin300450 China

While pioneering deep learning methods have made great strides in analyzing electronic health record (EHR) data, they often struggle to fully capture the semantics of diverse medical codes from limited data. The integration of external knowledge from Large Language Models (LLMs) presents a promising avenue for improving healthcare predictions. However, LLM analyses may exhibit significant variance due to ambiguity problems and inconsistency issues, hindering their effective utilization. To address these challenges, we propose IntelliCare, a novel framework that leverages LLMs to provide high-quality patient-level external knowledge and enhance existing EHR models. Concretely, IntelliCare identifies patient cohorts and employs task-relevant statistical information to augment LLM understanding and generation, effectively mitigating the ambiguity problem. Additionally, it refines LLM-derived knowledge through a hybrid approach, generating multiple analyses and calibrating them using both the EHR model and perplexity measures. Experimental evaluations on three clinical prediction tasks across two large-scale EHR datasets demonstrate that IntelliCare delivers significant performance improvements to existing methods, highlighting its potential in advancing personalized healthcare predictions and decision support systems. Our code is available at https://***/yzhHoward/IntelliCare. © 2025, The Authors. All rights reserved.

关键词： Electronic health record

来源：评论

学校读者我要写书评

暂无评论

Enhancing Medical Image-Based Diagnostics Through the Application of Convolutional Neural Networks Techniques

Enhancing Medical Image-Based Diagnostics Through the Applic...

引用

Distributed Computing and Electrical Circuits and Electronics (ICDCECE), IEEE International Conference on

作者： Haewon Byeon Rajit Nair V. Mahalakshmi Mohammed I. Khalaf Bharat Kaushik Mohammad Shabaz Department of AI and Software Inje University Gimhae Republic of Korea School of Computing Science & Engineering VIT Bhopal University Bhopal India Department of Computer Science College of Engineering and computer science Jazan University Jazan Saudi Arabia Department of computer science Al Maarif University College Al Anbar Iraq Masters University of Northern Iowa Lowa USA Model Institute of Engineering and Technology Jammu J&K India

ISBN: (数字)9798350318609

ISBN: (纸本)9798350318616

Perioperative nutrition evaluation is crucial for patient care and resource efficiency in healthcare. This work proposes a deep learning-based perioperative nutrition evaluation tool that leverages sophisticated machine learning to improve predictions and assist clinicians in making better choices. As surgical techniques become more complex and medical technology progresses, a thorough postoperative nutrition study becomes more important than ever. An extensive review demonstrates that the recommended strategy performs better than other machine learning techniques. Our model's accuracy, precision, memory, and F1-score let it predict surgical outcomes more accurately. The higher AVC-ROC and AVC-PR values of the suggested approach show that it performs particularly well. Its sensitivity-specificity balance is demonstrated by this. Our approach generates fewer errors, according to the confusion matrix analysis, which makes it reliable for healthcare decisions. Our work enhances critical healthcare by assessing nutrition with state-of-the-art techniques.

关键词： Fluctuations Reviews Hospitals Measurement uncertainty Surgery Medical services Predictive models

来源：评论

学校读者我要写书评

暂无评论

Revisit Raft Consistency Protocol on Private Blockchain System in High Network Latency 7th

Revisit Raft Consistency Protocol on Private Blockchain Syst...

引用

7th International Conference on Artificial Intelligence and Security, ICAIS 2021

作者： Cao, Ning Jiang, Dianheng Liu, Yang Zhou, Yulan Du, Haiwen Qiao, Xueming Xia, Yingxue Zhu, Dongjie Yu, Fang Bi, Wenbin College of Information Engineering Sanming University Sanming365000 China School of Computer Science and Technology Harbin Institute of Technology Weihai264209 China School of Astronautics Harbin Institute of Technology Harbin150001 China State Grid Weihai Power Supply Company No. 23 Kunming Road Weihai264200 China College of Information Engineering Qingdao Binhai University Qingdao266555 China Department of Internet of Things Engineering School of Computer and Software Dalian Neusoft University of Information Dalian116033 China

ISBN: (纸本)9783030786175

Raft’s good performance in the CFT system makes it a mainstream implementation solution for the consistency of the private chain system. However, we found in practice that Raft’s performance degradation is very serious when the network latency is greater than 1 ms. Therefore, we simulated the blockchain network environment and tested the TPS performance of Raft and Multi-raft. The results show that when the TPS is the same, Multi-raft reduces network traffic by 30% compared with Raft’s leader, and the CPU load increases slightly. Based on the experimental results, we discussed the optimization scheme of Raft on the geographically distributed system, and provided ideas for future research. © 2021, Springer Nature Switzerland AG.

关键词： Blockchain

来源：评论

学校读者我要写书评

暂无评论

Through the Sci-Fi Lens: Exploring Social Media’s Influence on Body Image and Bullying Among Adolescents in Boarding schools

Through the Sci-Fi Lens: Exploring Social Media’s Influence...

引用

Learning and technology Conference (L&T)

作者： Jawahir Che Mustapha Muhammad Azmin Mohamed Ghazali Munaisyah Abdullah Nor Intan Saniah Sulaiman Hezlin Aryani Abd Rahman Informatics & Analytics Section Universiti Kuala Lumpur Malaysian Institute of Information Technology Kuala Lumpur Malaysia Computer Engineering Tech. Section Universiti Kuala Lumpur Malaysian Institute of Information Technology Kuala Lumpur Malaysia Software Engineering Section Universiti Kuala Lumpur Malaysian Institute of Information Technology Kuala Lumpur Malaysia College of Arts and Science School of Quantitative Sciences Universiti Utara Malaysia Kedah Malaysia Centre of Statistical and Decision Science Studies Faculty of Computer and Mathematical Sciences Universiti Teknologi MARA Selangor Malaysia

This research investigates the complex relationships between social media, body image, and bullying among adolescent boarding school students, with an emphasis on the Malaysian context. Utilizing science fiction as a conceptual lens, the study presents hypothetical future scenarios to investigate how technological advancements could impact these interconnected issues. The paper seeks to provide nuanced insights into the ethical and policy implications of social media use in boarding schools by drawing on key theories such as Social Comparison Theory, Social-Ecological Model and Uses and Gratifications Theory. Through comparative analysis of hypothetical scenarios, this study aims to provide educators, policymakers, and technologists with a comprehensive understanding of the challenges and opportunities at this intersection. The ultimate objective is to contribute to the ongoing discourse concerning adolescent mental health and to propose actionable strategies for mitigating negative outcomes in the unique context of Malaysian boarding schools.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Rethinking Client Drift in Federated Learning: A Logit Perspective

arXiv

引用

arXiv 2023年

作者： Yan, Yunlu Feng, Chun-Mei Ye, Mang Zuo, Wangmeng Li, Ping Goh, Rick Siow Mong Zhu, Lei Chen, C.L. Philip Guangzhou Nansha511400 China The Institute of High Performance Computing A*STAR Singapore138632 Singapore The Hubei Luojia Laboratory National Engineering Research Center for Multimedia Software School of Computer Science Wuhan University Wuhan430072 China The School of Computer Science and Technology Harbin Institute of Technology Harbin130407 China The Hong Kong University of Science and Technology Hong Kong The Department of Computing the School of Design The Hong Kong Polytechnic University Hong Kong The School of Computer Science and Engineering South China University of Technology Guangzhou510006 China The Pazhou Lab Guangzhou510335 China

Federated Learning (FL) enables multiple clients to collaboratively learn in a distributed way, allowing for privacy protection. However, the real-world non-IID data will lead to client drift which degrades the performance of FL. Interestingly, we find that the difference in logits between the local and global models increases as the model is continuously updated, thus seriously deteriorating FL performance. This is mainly due to catastrophic forgetting caused by data heterogeneity between clients. To alleviate this problem, we propose a new algorithm, named FedCSD, a Class prototype Similarity Distillation in a federated framework to align the local and global models. FedCSD does not simply transfer global knowledge to local clients, as an undertrained global model cannot provide reliable knowledge, i.e., class similarity information, and its wrong soft labels will mislead the optimization of local models. Concretely, FedCSD introduces a class prototype similarity distillation to align the local logits with the refined global logits that are weighted by the similarity between local logits and the global prototype. To enhance the quality of global logits, FedCSD adopts an adaptive mask to filter out the terrible soft labels of the global models, thereby preventing them to mislead local optimization. Extensive experiments demonstrate the superiority of our method over the state-of-the-art federated learning approaches in various heterogeneous settings. The source code will be released. Copyright © 2023, The Authors. All rights reserved.

关键词： Distillation

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：