检索结果-内蒙古大学图书馆

A prompt-based approach to adversarial example generation and robustness enhancement

Frontiers of computer science 2024年第4期18卷 85-96页

作者： Yuting YANG Pei HUANG Juan CAO Jintao LI Yun LIN Feifei MA Key Lab of Intelligent Information Processing of Chinese Academy of Sciences(CAS) Institute of Computing TechnologyCASBeijing 100190China School of Computer Science and Technology University of Chinese Academy of SciencesBeijing 100049China Department of Computer Science Stanford UniversityCA 94305USA School of Computing National University of SingaporeSingapore 119077Singapore Laboratory of Parallel Software and Computational Science Institute of SoftwareChinese Academy of SciencesBeijing 100190China

Recent years have seen the wide application of natural language processing(NLP)models in crucial areas such as finance,medical treatment,and news media,raising concerns about the model robustness and *** find that prompt paradigm can probe special robust defects of pre-trained language *** prompt texts are first constructed for inputs and a pre-trained language model can generate adversarial examples for victim models via *** results show that prompt paradigm can efficiently generate more diverse adversarial examples besides synonym ***,we propose a novel robust training approach based on prompt paradigm which incorporates prompt texts as the alternatives to adversarial examples and enhances robustness under a lightweight minimax-style optimization *** on three real-world tasks and two deep neural models show that our approach can significantly improve the robustness of models to resist adversarial attacks.

关键词： robustness adversarial example prompt learning pre-trained language model

来源：评论

学校读者我要写书评

暂无评论

Unsupervised Deep Learning: Taxonomy and Algorithms

引用

Informatica (Slovenia) 2022年第2期46卷 151-168页

作者： Chefrour, Aida Souici-Meslati, Labiba Computer Science Department Mohamed Cherif Messaadia University Souk Ahras Algeria LISCO Laboratory Computer Science Department Badji Mokhtar University B.P-12 Annaba23000 Algeria

Clustering is a fundamental challenge in many data-driven application fields and machine learning techniques. The data distribution determines the quality of the outcomes, which has a significant impact on clustering performance. As a result, deep neural networks can be used to learn more accurate data representations for clustering. Many recent studies have focused on employing deep neural networks to develop a clustering-friendly representation, which has resulted in a significant improvement in clustering performance. We present a systematic survey of clustering with deep learning in this study. Then, a taxonomy of deep clustering is proposed, as well as some sample algorithms for our overview. Finally, we discuss some exciting future possibilities for clustering using deep learning and offer some remarks. © 2022 Slovene Society Informatika. All rights reserved.

关键词： Taxonomies

来源：评论

学校读者我要写书评

暂无评论

Unlocking Music Archives: Openness and Accessibility 21

Unlocking Music Archives: Openness and Accessibility

引用

21st Conference on Information and Research science Connecting to Digital and Library science, IRCDL 2025

作者： Faschi, Vanessa Avanzini, Federico Ludovico, Luca Andrea Laboratory of Music Informatics Department of Computer Science University of Milan via G. Celoria 18 Milan20135 Italy

This paper focuses on the theme of unlocking the materials preserved by musical institutions such as opera houses, conservatories, and historical archives. In this context, "unlocking" means providing access to anyone, which can be interpreted in two different ways. The first meaning, related to openness, implies making preserved materials accessible to a wide audience. From this perspective, digitization campaigns and the subsequent enhancement of cultural assets through technology can open the way to new possibilities. For example, creating a dedicated Web portal can facilitate access by addressing issues such as physical distances, availability at any time of the day, the perishability of physical objects, insurance costs, etc. Moreover, technology can provide new services and foster advanced ways to exploit materials. Concerning the second meaning, related to accessibility, technologies can be used to assist people with physical and cognitive impairments and extend the use of these materials to people for whom it would be more difficult. For example, assistive technologies, such as screen readers or braille displays, allow blind and visually impaired (BVI) users to enjoy the transcription of music or literary texts. Similarly, computer-assisted approaches, such as large language models, can reduce the complexity of a text to meet the communicative abilities of people with special educational needs or cognitive impairments. To better illustrate the meanings of openness and accessibility, we will discuss two scenarios excerpted from the Ricordi Historical Archive. On the one hand, this work aims to highlight the fundamental role of technologies in supporting openness and accessibility, starting from digitization campaigns, through the online publishing of materials, to the release of ad hoc software tools. On the other hand, our research shows that there is still a long way to go before music archives can fully achieve the goals of openness and accessibility. © 20

关键词： Accessibility Archives Collections Inclusivity Music Usability

来源：评论

学校读者我要写书评

暂无评论

An Interpretable CNN for the Segmentation of the Left Ventricle in Cardiac MRI by Real-Time Visualization

引用

computer Modeling in Engineering & sciences 2023年第5期135卷 1571-1587页

作者： Jun Liu Geng Yuan Changdi Yang Houbing Song Liang Luo Robotics Institute School of Computer ScienceCarnegie Mellon UniversityPittsburghPA15217USA Department of Electrical&Computer Engineering College of EngineeringNortheastern UniversityBostonMA02115USA Security and Optimization for Networked Globe Laboratory(SONG Lab) Embry-Riddle Aeronautical UniversityDaytona BeachFL32114USA School of Computer Science and Engineering University of Electronic Science and Technology of ChinaChengdu610054China

The interpretability of deep learning models has emerged as a compelling area in artificial intelligence *** safety criteria for medical imaging are highly stringent,and models are required for an ***,existing convolutional neural network solutions for left ventricular segmentation are viewed in terms of inputs and ***,the interpretability of CNNs has come into the *** medical imaging data are limited,many methods to fine-tune medical imaging models that are popular in transfer models have been built using massive public Image Net datasets by the transfer learning ***,this generates many unreliable parameters and makes it difficult to generate plausible explanations from these *** this study,we trained from scratch rather than relying on transfer learning,creating a novel interpretable approach for autonomously segmenting the left ventricle with a cardiac *** enhanced GPU training system implemented interpretable global average pooling for graphics using deep *** deep learning tasks were *** included data management,neural network architecture,and *** system monitored and analyzed the gradient changes of different layers with dynamic visualizations in real-time and selected the optimal deployment *** results demonstrated that the proposed method was feasible and efficient:the Dice coefficient reached 94.48%,and the accuracy reached 99.7%.It was found that no current transfer learning models could perform comparably to the ImageNet transfer learning *** model is lightweight and more convenient to deploy on mobile devices than transfer learning models.

关键词： Interpretable graphics training visualization image segmentation left ventricle CNNs global average pooling

来源：评论

学校读者我要写书评

暂无评论

A Simple but Effective Approach for Unsupervised Few-Shot Graph Classification 24

A Simple but Effective Approach for Unsupervised Few-Shot Gr...

引用

33rd ACM Web Conference, WWW 2024

作者： Liu, Yonghao Huang, Lan Cao, Bowen Li, Ximing Giunchiglia, Fausto Feng, Xiaoyue Guan, Renchu College of Computer Science and Technology Jilin University Changchun China Department of Information Engineering and Computer Science University of Trento Trento Italy Key Laboratory of Symbolic Computation and Knowledge Engineering of the Ministry of Education China

ISBN: (纸本)9798400701719

Graphs, as a fundamental data structure, have proven efficacy in modeling complex relationships between objects and are therefore found in wide web applications. Graph classification is an essential task in graph data analysis, which can effectively assist in extracting information and mining content from the web. Recently, few-shot graph classification, a more realistic and challenging task, has garnered great research interest. Existing few-shot graph classification models are all supervised, assuming abundant labeled data in base classes for meta-training. However, sufficient annotation is often challenging to obtain in practice due to high costs or demand for expertise. Moreover, they commonly adopt complicated meta-learning algorithms via episodic training to transfer prior knowledge from base classes. To break free from these constraints, in this paper, we propose a simple yet effective approach named SMART for unsupervised few-shot graph classification without using any labeled data. SMART employs transfer learning philosophy instead of the previously prevailing meta-learning paradigm, avoiding the need for sophisticated meta-learning algorithms. Additionally, we adopt a novel mixup strategy to augment the original graph data and leverage unsupervised pretraining on these data to obtain the expressive graph encoder. We also utilize the prompt tuning technique to alleviate the overfitting and low fine-tuning efficiency caused by the limited support samples of novel classes. Extensive experimental results demonstrate the superiority of our proposed approach, significantly surpassing even leading supervised few-shot graph classification models. Our code is available here. © 2024 ACM.

关键词： Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

PierGuard: A Planning Framework for Underwater Robotic Inspection of Coastal Piers

引用

IEEE Transactions on Automation science and Engineering 2025年 22卷 15941-15952页

作者： Wang, Pengyu Lin, Hin Wang Li, Jialu Wang, Jiankun Shi, Ling Meng, Max Q.-H Southern University of Science and Technology Shenzhen Key Laboratory of Robotics Perception and Intelligence Department of Electronic and Electrical Engineering Shenzhen China Hong Kong University of Science and Technology Department of Electronic and Computer Engineering Hong Kong Southern University of Science and Technology Jiaxing Research Institute Jiaxing China Hong Kong University of Science and Technology Department of Chemical and Biological Engineering Hong Kong The Chinese University of Hong Kong Department of Electronic Engineering Hong Kong University of Alberta Canada Department of Electrical and Computer Engineering Canada

Using underwater robots instead of humans for the inspection of coastal piers can enhance efficiency while reducing risks. A key challenge in performing these tasks lies in achieving efficient and rapid path planning within complex environments. Sampling-based path planning methods, such as Rapidly-exploring Random Tree* (RRT*), have demonstrated notable performance in high-dimensional spaces. In recent years, researchers have begun designing various geometry-inspired heuristics and neural network-driven heuristics to further enhance the effectiveness of RRT*. However, the performance of these general path planning methods still requires improvement when applied to highly cluttered underwater environments. In this paper, we propose PierGuard, which combines the strengths of bidirectional search and neural network-driven heuristic regions. We design a specialized neural network to generate high-quality heuristic regions in cluttered maps, thereby improving the performance of the path planning. Through extensive simulation and real-world ocean field experiments, we demonstrate the effectiveness and efficiency of our proposed method compared with previous research. Our method achieves approximately 2.6 times the performance of the state-of-the-art geometric-based sampling method and nearly 4.9 times that of the state-of-the-art learning-based sampling method. Our results provide valuable insights for the automation of pier inspection and the enhancement of maritime safety. (Video1). © 2004-2012 IEEE.

关键词： Motion planning

来源：评论

学校读者我要写书评

暂无评论

DFA-GNN: Forward Learning of Graph Neural Networks by Direct Feedback Alignment 38

DFA-GNN: Forward Learning of Graph Neural Networks by Direct...

引用

38th Conference on Neural Information Processing Systems, NeurIPS 2024

作者： Zhao, Gongpei Wang, Tao Lang, Congyan Jin, Yi Li, Yidong Ling, Haibin Key Laboratory of Big Data & Artificial Intelligence in Transportation Ministry of Education China School of Computer Science & Technology Beijing Jiaotong University China Department of Computer Science Stony Brook University United States

Graph neural networks (GNNs) are recognized for their strong performance across various applications, with the backpropagation (BP) algorithm playing a central role in the development of most GNN models. However, despite its effectiveness, BP has limitations that challenge its biological plausibility and affect the efficiency, scalability and parallelism of training neural networks for graph-based tasks. While several non-backpropagation (non-BP) training algorithms, such as the direct feedback alignment (DFA), have been successfully applied to fully-connected and convolutional network components for handling Euclidean data, directly adapting these non-BP frameworks to manage non-Euclidean graph data in GNN models presents significant challenges. These challenges primarily arise from the violation of the independent and identically distributed (i.i.d.) assumption in graph data and the difficulty in accessing prediction errors for all samples (nodes) within the graph. To overcome these obstacles, in this paper we propose DFA-GNN, a novel forward learning framework tailored for GNNs with a case study of semi-supervised learning. The proposed method breaks the limitations of BP by using a dedicated forward training mechanism. Specifically, DFA-GNN extends the principles of DFA to adapt to graph data and unique architecture of GNNs, which incorporates the information of graph topology into the feedback links to accommodate the non-Euclidean characteristics of graph data. Additionally, for semi-supervised graph learning tasks, we developed a pseudo error generator that spreads residual errors from training data to create a pseudo error for each unlabeled node. These pseudo errors are then utilized to train GNNs using DFA. Extensive experiments on 10 public benchmarks reveal that our learning framework outperforms not only previous non-BP methods but also the standard BP methods, and it exhibits excellent robustness against various types of noise and attacks. © 2024 Neura

关键词：

来源：评论

学校读者我要写书评

暂无评论

Deep Classifier Fusion-based Multi-Classification of Chest X-rays for COVID-19 Detection

Deep Classifier Fusion-based Multi-Classification of Chest X...

引用

2023 International Conference on Decision Aid sciences and Applications, DASA 2023

作者： Benzebouchi, Nacer Eddine Bourzama, Mohamed Ouassim Zemmal, Nawel Azizi, Nabiha Badji Mokhtar Annaba University Labged Laboratory Computer Science Department Annaba23000 Algeria Badji Mokhtar Annaba University Computer Science Department Annaba23000 Algeria Mohamed Cherif Messaadia University Mathematics and Computer Science Department Souk-Ahras Algeria

ISBN: (纸本)9798350342055

In the last three years, The COVID-19 pandemic has placed immense strain on healthcare systems worldwide. Imaging techniques, such as Chest X-Ray (CXR) images, play an essential role in diagnosing COVID-19. Recently, computer Aided Diagnosis systems based on Deep Learning have been widely used to identify COVID-19 from other upper respiratory diseases such as viral pneumonia. To enhance COVID-19 diagnosis using CXR images, this article proposes a methodology that combines three deep learning models renowned for their depth and feature extraction capabilities: ResNet, DenseNet, and VGG. The model' strengths are harnessed using Aggregation techniques such as stacking and voting (majority, weighted, average), to improve prediction accuracy. Moreover, the used dataset was subjected to preprocessing, balancing, and augmentation techniques, including rotation and zooming, to ensure robust recognition of CXR images, even in challenging positions. Experimental results demonstrate that the combination of model ensembles and enhanced dataset leads to superior accuracy in diagnosing COVID-19, with an impressive 97.95% accuracy, 98.2% F1 score, and 99.9% AUC-ROC. This approach not only outperforms individual models but also offers a more reliable and efficient solution for COVID-19 detection, particularly in complex cases where symptoms overlap with other respiratory illnesses. © 2023 IEEE.

关键词： COVID-19

来源：评论

学校读者我要写书评

暂无评论

Online Matching based Firefly Algorithm for Dynamic Ridesharing 13

Online Matching based Firefly Algorithm for Dynamic Rideshar...

引用

13th International Conference on Research in ComputIng at Feminine, RIF 2024

作者： Abdelmoumène, Hiba Bousena, Amina Computer Science Department University of 8 Mai 1945 Guelma Algeria LabGED Laboratory University of Badji Mokhtar Annaba Algeria

Urban transportation has been a significant contributor to environmental challenges, including traffic congestion and greenhouse gas emissions, exacerbating climate change. In contrast, public transportation, especially in rural areas, often struggles to meet mobility needs effectively. Ridesharing systems have emerged as a response to these pressing issues. By facilitating the sharing of rides among individuals heading in the same direction, ridesharing optimizes vehicle use. This not only offers economic benefits but also addresses environmental concerns. This study addresses the critical challenge of dynamic matching in ridesharing systems, where the objective is to optimize the allocation of drivers to multiple passengers. Recognizing the multifaceted nature of this problem, our approach leverages the Firefly algorithm, a metaheuristic known for its efficacy in tackling complex optimization tasks. In our model, we account for a range of constraints including spatiotemporal limitations, capacity consideration and passenger waiting time restrictions. The primary goal is to minimize both the waiting time for passengers and the total distance traveled. To rigorously evaluate our method, we conducted experiments utilizing real-world data from a geographic area in the city of Guelma, Algeria. The results obtained through our experiments have demonstrated the effectiveness and performance of our system based on the Firefly algorithm. Indeed, our approach enhances the experience of passengers via a notable reduction in waiting times and it contributes to the overall efficiency of the ridesharing system by minimizing the total distance traveled. © 2024 Copyright for this paper by its authors.

关键词： Traffic congestion

来源：评论

学校读者我要写书评

暂无评论

A NLP and Rule-Based Approach to Extract Spatial Entities and Relationships in Arabic Text 7

A NLP and Rule-Based Approach to Extract Spatial Entities an...

引用

7th International Conference on Informatics and Applied Mathematics, IAM 2024

作者： Hadji, Atmane Kholladi, Mohamed Khireddine Boumaza, Farid University of Bejaia Faculty of exact sciences Department of Computer Science Bejaia06000 Algeria LISI Laboratory Computer Science Department University Center A. Boussouf Mila Mila43000 Algeria Department of Mathematics and Computer Science Faculty of Exact Sciences University of El-Oued HAMMA Lakhdar El-Oued Algeria MISC Laboratory of Abdelhamid Mehri university of Constantine 2 Algeria Computer Science Department University of Mohamed El Bachir El Ibrahimi Bordj Bou Arreridj34030 Algeria LAPECI Laboratory University of Oran1 Oran31000 Algeria

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：