检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Hou, Bohan Lin, Haoqiang Wen, Haokun Liu, Meng Xu, Mingzhu Song, Xuemeng The Department of Computer Science and Technology Shandong University Qingdao China Shenzhen China The Department of Data Science City University of Hong Kong Hong Kong The Department of Computer Science Shandong Jianzhu University Jinan China The School of Software Shandong University Jinan China

Composed Image Retrieval (CIR) is a challenging task that aims to retrieve the target image with a multimodal query, i.e., a reference image, and its complementary modification text. As previous supervised or zero-shot learning paradigms all fail to strike a good trade-off between the model’s generalization ability and retrieval performance, recent researchers have introduced the task of few-shot CIR (FS-CIR) and proposed a textual inversion-based network based on pretrained CLIP model to realize it. Despite its promising performance, the approach encounters two key limitations: simply relying on the few annotated samples for CIR model training and indiscriminately selecting training triplets for CIR model fine-tuning. To address these two limitations, we propose a novel two-stage pseudo triplet guided few-shot CIR scheme, dubbed PTG-FSCIR. In the first stage, we propose an attentive masking and captioning-based pseudo triplet generation method, to construct pseudo triplets from pure image data and use them to fulfill the CIR-task specific pretraining. In the second stage, we propose a challenging triplet-based CIR fine-tuning method, where we design a pseudo modification text-based sample challenging score estimation strategy and a robust top range-based random sampling strategy for sampling robust challenging triplets to promote the model fine-tuning. Notably, our scheme is plug-and-play and compatible with any existing supervised CIR models. We test our scheme across two backbones on three public datasets (i.e., FashionIQ, CIRR, and Birds-to-Words), achieving maximum improvements of 13.3%, 22.2%, and 17.4% respectively, demonstrating our scheme’s efficacy. © 2024, CC BY.

关键词： Zero-shot learning

来源：评论

学校读者我要写书评

暂无评论

Exploring Reinforcement Learning in Large-Scale data Processing 3

Exploring Reinforcement Learning in Large-Scale Data Process...

引用

3rd International Conference on Smart Generation Computing, Communication and Networking, SMART GENCON 2023

作者： Bhambu, Pawan Kumar, Ritesh Sharmila, P. Patil, Vivek D. Khurana, Shivam Vivek, V. Vivekananda Global University Department of Computer Science & Engineering Jaipur India Maharishi School of Engineering and Technology Maharishi University of Information Technology Uttar Pradesh India Prince Shri Venkateshwara Padmavathy Engineering College Department of Science and Humanities Chennai127 India Vishwakarma Institute of Information Technology Department of Artificial Intelligence & Data Science Pune India Chitkara University Institute of Engineering and Technology Chitkara University Centre of Interdisciplinary Research in Business and Technology Punjab India Computer Science and Engineering Faculty of Engineering and Technology Karnataka Bangalore India

ISBN: (纸本)9798350319125

Reinforcement mastering (RL) is vital for big-scale fast processing. RL utilizes the availability of large datasets and dealers skilled in this information to remedy a wide variety of obligations, from automated diagnostics and self-reliant using structures to clinical information analysis. Using exploiting the traits of the modern-day significant information era, RL allows efficient and correct decision-making in massive-scale information processing and answers complicated troubles. This paper explores the use of reinforcement learning (RL) in large-scale data processing. RL is a subfield of machine learning that focuses on decision making in dynamic environments. With the increasing demand for processing large amounts of data, traditional techniques such as rule-based systems and batch processing are becoming inadequate. This has led to the emergence of RL as a potential solution due to its ability to learn and adapt to changing environments. Through a comprehensive review of existing literature, this paper discusses the applications of RL in large-scale data processing and its potential benefits. Furthermore, it examines various challenges and their potential solutions in implementing RL for data processing tasks. The paper concludes by highlighting the future research directions in this domain and the potential impact of RL on improving data processing efficiency and accuracy. It similarly examines some of the demanding situations and possibilities related to using RL for huge-scale records processing and the capability programs of RL in the future. © 2023 IEEE.

关键词： applications challenges opportunities potential processing

来源：评论

学校读者我要写书评

暂无评论

Decision Systems in Cybersecurity: A Machine Learning Perspective 12th

Decision Systems in Cybersecurity: A Machine Learning Perspe...

引用

12th International Conference on Frontiers of Intelligent Computing: Theory and Applications, FICTA 2024

作者： Dhablia, Anishkumar Ganesh, R. Senthil Rizvi, Nuzhat Arora, Himanshu Borade, Monali Ravindra Bhujbal, Pallavi Altimetrik India Pvt. Ltd Maharashtra Pune India Department of Electronics and Communication Engineering Sri Krishna College of Engineering and Technology Tamilnadu Coimbatore India Nagpur Campus Pune India Computer Science and Engineering Arya College of Engineering Jaipur India Artificial Intelligence and Data Science VIIT Pune India Department of Information Technology School of Computing MIT ADT University Loni Kalbhor Pune India

ISBN: (纸本)9789819601462

The evolving digital environment requires sophisticated decision systems to protect important information assets from the growing complexity and variety of cyber threats. This review paper examines the impact of machine learning (ML) on cybersecurity decision systems, with a specific focus on analyzing ML and deep learning (DL) methods. The study employs the UNSW-NB15 dataset, which is a widely used benchmark dataset in the field of network security. The paper starts with an overview of the present cybersecurity environment, emphasizing the difficulties presented by advanced and changing cyber threats. It highlights the importance of adaptive and intelligent decision systems that can efficiently identify and reduce cyberattacks as they occur. A significant part of the paper focuses on a thorough analysis of different machine learning and deep learning techniques used in cybersecurity. The review discusses conventional machine learning algorithms like decision trees, support vector machines, and ensemble methods, along with sophisticated deep learning architectures such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs). Each approach is thoroughly assessed for its strengths and limitations based on factors like accuracy, interpretability, and scalability. The performance of various ML and DL models is evaluated using the UNSW-NB15 dataset as a benchmark. The dataset covers various cyberattack scenarios, enabling a thorough assessment of algorithms’ ability to detect anomalies and classify malicious activities. The paper explores the incorporation of machine learning (ML) and deep learning (DL) into decision support systems, highlighting the significance of explainability and interpretability in the realm of cybersecurity. The conversation covers the difficulties of implementing machine learning (ML) and deep learning (DL) models in practical situations, such as concerns about data privacy, adversarial attacks, and model resilience. This rev

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Entropic Gromov-Wasserstein distances: stability and algorithms

The Journal of Machine Learning Research

引用

The Journal of Machine Learning Research 2024年第1期25卷 17738-17789页

作者： Gabriel Rioux Ziv Goldfeld Kengo Kato Center for Applied Mathematics Cornell University Ithaca NY School of Electrical and Computer Engineering Cornell University Ithaca NY Department of Statistics and Data Science Cornell University Ithaca NY

The Gromov-Wasserstein (GW) distance quantifies discrepancy between metric measure spaces and provides a natural framework for aligning heterogeneous datasets. Alas, as exact computation of GW alignment is NP-complete, entropic regularization provides an avenue towards a computationally tractable proxy. Leveraging a recently derived variational representation for the quadratic entropic GW (EGW) distance, this work derives the first efficient algorithms for solving the EGW problem subject to formal, nonasymptotic convergence guarantees. To that end, we derive smoothness and convexity properties of the objective in this variational problem, which enables its resolution by the accelerated gradient method. Our algorithms employ Sinkhorn's fixed point iterations to compute an approximate gradient, which we model as an inexact oracle. We furnish convergence rates towards local and even global solutions (the latter holds under a precise quantitative condition on the regularization parameter), characterize the effects of gradient inexactness, and prove that stationary points of the EGW problem converge towards a stationary point of the unregularized GW problem, in the limit of vanishing regularization. We provide numerical experiments that validate our theory and empirically demonstrate the state-of-the-art empirical performance of our algorithm.

关键词： Algorithms convergence rate global guarantees Gromov-Wasserstein distances entropic regularization inexact gradient methods

来源：评论

学校读者我要写书评

暂无评论

Visual Evaluative AI: A Hypothesis-Driven Tool with Concept-Based Explanations and Weight of Evidence

arXiv

引用

arXiv 2024年

作者： Le, Thao Miller, Tim Zhang, Ruihan Sonenberg, Liz Singh, Ronal School of Computing and Information Systems The University of Melbourne Australia School of Electrical Engineering and Computer Science The University of Queensland Australia CSIRO/Data61 Australia

This paper presents Visual Evaluative AI, a decision aid that provides positive and negative evidence from image data for a given hypothesis. This tool finds high-level human concepts in an image and generates the Weight of Evidence (WoE) for each hypothesis in the decision-making process. We apply and evaluate this tool in the skin cancer domain by building a web-based application that allows users to upload a dermatoscopic image, select a hypothesis and analyse their decisions by evaluating the provided evidence. Further, we demonstrate the effectiveness of Visual Evaluative AI on different concept-based explanation approaches. Copyright © 2024, The Authors. All rights reserved.

关键词： Decision making

来源：评论

学校读者我要写书评

暂无评论

Mixed Attention Densely Residual Network for Single Image Super-Resolution

引用

computer Systems science & Engineering 2021年第10期39卷 133-146页

作者： Jingjun Zhou Jing Liu Jingbing Li Mengxing Huang Jieren Cheng Yen-Wei Chen Yingying Xu Saqib Ali Nawaz School of Information and Communication Engineering Hainan UniversityHaikou570228China State Key Laboratory of Marine Resource Utilization in the South China Sea Hainan UniversityHaikou570228China Research Center for Healthcare Data Science Zhejiang LabHangzhou311121China School of Computer Science and Cyberspace Security Hainan UniversityHaikou570228China Graduate School of Information Science and Engineering Ritsumeikan University5258577Japan College of Computer Science and Technology Zhejiang UniversityHangzhou311100China

Recent applications of convolutional neural networks(CNNs)in single image super-resolution(SISR)have achieved unprecedented ***,existing CNN-based SISR network structure design consider mostly only channel or spatial information,and cannot make full use of both channel and spatial information to improve SISR performance *** present work addresses this problem by proposing a mixed attention densely residual network architecture that can make full and simultaneous use of both channel and spatial ***,we propose a residual in dense network structure composed of dense connections between multiple dense residual groups to form a very deep *** structure allows each dense residual group to apply a local residual skip connection and enables the cascading of multiple residual blocks to reuse previous features.A mixed attention module is inserted into each dense residual group,to enable the algorithm to fuse channel attention with laplacian spatial attention effectively,and thereby more adaptively focus on valuable feature *** qualitative and quantitative results of extensive experiments have demonstrate that the proposed method has a comparable performance with other stateof-the-art methods.

关键词： Channel attention Laplacian spatial attention residual in dense mixed attention

来源：评论

学校读者我要写书评

暂无评论

EdgeOAR: Real-time Online Action Recognition On Edge Devices

arXiv

引用

arXiv 2024年

作者： Luo, Wei Zhang, Deyu Tang, Ying Wu, Fan Zhang, Yaoxue School of Computer Science and Engineering Central South University Changsha410083 China Big Data Institute Central South University Changsha410083 China Department of Computer Science and Technology Tsinghua University Beijing100084 China

This paper addresses the challenges of Online Action Recognition (OAR), a framework that involves instantaneous analysis and classification of behaviors in video streams. OAR must operate under stringent latency constraints, making it an indispensable component for real-time feedback for edge computing. Existing methods, which typically rely on the processing of entire video clips, fall short in scenarios requiring immediate recognition. To address this, we designed EdgeOAR, a novel framework specifically designed for OAR on edge devices. EdgeOAR includes the Early Exit-oriented Task-specific Feature Enhancement Module (TFEM), which comprises lightweight submodules to optimize features in both temporal and spatial dimensions. We design an iterative training method to enable TFEM learning features from the beginning of the video. Additionally, EdgeOAR includes an Inverse Information Entropy (IIE) and Modality Consistency (MC)-driven fusion module to fuse features and make better exit decisions. This design overcomes the two main challenges: robust modeling of spatio-temporal action representations with limited initial frames in online video streams and balancing accuracy and efficiency on resource-constrained edge devices. Experiments show that on the UCF-101 dataset, our method EdgeOAR reduces latency by 99.23% and energy consumption by 99.28% compared to state-of-the-art (SOTA) method. And achieves an adequate accuracy on edge devices. © 2024, CC BY.

关键词： Spatio-temporal data

来源：评论

学校读者我要写书评

暂无评论

Channel-wise attention based binarization for CNN Architectures 5

Channel-wise attention based binarization for CNN Architectu...

引用

5th International Conference on Emerging Systems and Intelligent Computing, ESIC 2025

作者： Paul, Ipsita Bandyopadhyay, Mainak Dash, Satya Ranjan Aurpa, Tanjim Taharat Kiit D.U. School of Computer Engineering Bhubaneswar India Kiit D.U. School of Computer Applications Bhubaneswar India Bsmr Digital University Department of Data Science Engineering Bangladesh

ISBN: (纸本)9798331522100

Attention mechanisms are known to effectively increase the performance of a model by focusing on specific sections of the input while executing a task. Squeeze and Excitation module (SE-Net) and Efficient channel attention module (ECA-Net) are well-known Channel Attention networks that only involve a few parameters to achieve a clear performance gain. In order to further improve the model's accuracy, Binarized Neural Networks (BNNs) can be introduced in the same. BNNs are deep learning models that use binarized values for activation functions and training weights, instead of fully precised values that leads to much faster computation, occupying less memory and power. Motivated by this, a method of binarization of the convolutional network is proposed in this paper that can enhance the overall model accuracy. In our work, we focused on evaluating the performance of our binarized attention module over the pre-existing ones. Thus, the objective of this work is to design, implement, and evaluate efficient neural network models that balance accuracy and computational efficiency for image classification tasks on datasets with differing complexity levels, such as CIFAR-10 and CIFAR-100. By exploring various backbone architectures (e.g., MobileNetV2 and ResNet50), attention mechanisms (e.g., ECANet and SENet), and binarization techniques, this study aims to understand the impact of these components on model performance and resource requirements. Ultimately, the goal is to develop optimized models that achieve high accuracy with minimal computational overhead, making them suitable for deployment in resource-constrained environments. © 2025 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Recent Trends for Text Summarization in Scientific Documents

Recent Trends for Text Summarization in Scientific Documents

引用

International Conference on Computing, Engineering, and Design (ICCED)

作者： Arief Agus Sukmandhani Yulyani Arifin Muhammad Zarlis Widodo Budiharto Computer Science Department Binus Graduate Program Computer Science Bina Nusantara University Jakarta Indonesia Information Systems Management Department BINUS Graduate Program - Master of Information Systems Management Bina Nusantara University Jakarta Indonesia Computer Science Department School of Computer Science Bina Nusantara University Jakarta Indonesia

Text summarization is a natural language processing (NLP) technique in artificial intelligence that has been studied in recent years. Every document containing text is tested to get a good summary result. In producing a good summary, proper accuracy is required; therefore, various techniques and methods are used, and another challenge is the use of language in the documents being tested. Various types of documents have been used as research for text summarization, including scientific documents, so researchers are interested in studying them more intensely to get the state of the art in research. This research was conducted to review previous research using literature studies from various publications that published research on text summarization in scientific documents. The researcher collected data from several publishers who had published this type of research from 1988 to 2023 quarter 1 and obtained three hundred and eleven articles, then the researcher analyzed the data using the Kitchenham method to get the right literature to answer the research question. Thirty-seven articles were selected for further in-depth analysis. In addition, the researcher found some interesting findings beyond the research question, including other techniques used to solve the text summarization problem and other types of output that are not commonly produced so that it can increase knowledge in researching text summarization.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Resampling Factor Estimation via Dual-Stream Convolutional Neural Network

引用

computers, Materials & Continua 2021年第1期66卷 647-657页

作者： Shangjun Luo Junwei Luo Wei Lu Yanmei Fang Jinhua Zeng Shaopei Shi Yue Zhang School of Data and Computer Science Guangdong Province Key Laboratory of Information Security TechnologyMinistry of Education Key Laboratory of Machine Intelligence and Advanced ComputingSun Yat-sen UniversityGuangzhou510006China Academy of Forensic Science Shanghai200063China College of Information Science and Technology Jinan UniversityGuangzhou510632China Department of Computer Science University of Massachusetts LowellLowellMA 01854USA

The estimation of image resampling factors is an important problem in image *** all the resampling factor estimation methods,spectrumbased methods are one of the most widely used methods and have attracted a lot of research ***,because of inherent ambiguity,spectrum-based methods fail to discriminate upscale and downscale operations without any prior *** general,the application of resampling leaves detectable traces in both spatial domain and frequency domain of a resampled ***,the resampling process will introduce correlations between neighboring *** this case,a set of periodic pixels that are correlated to their neighbors can be found in a resampled ***,the resampled image has distinct and strong peaks on spectrum while the spectrum of original image has no clear ***,in this paper,we propose a dual-stream convolutional neural network for image resampling factors *** of the two streams is gray stream whose purpose is to extract resampling traces features directly from the rescaled *** other is frequency stream that discovers the differences of spectrum between rescaled and original *** features from two streams are then fused to construct a feature representation including the resampling traces left in spatial and frequency domain,which is later fed into softmax layer for resampling factor *** results show that the proposed method is effective on resampling factor estimation and outperforms some CNN-based methods.

关键词： Image forensics image resampling detection parameter estimation convolutional neural network

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：