检索结果-内蒙古大学图书馆

34th ACM Web Conference, WWW 2025

作者： Wang, Kaibin He, Qiang Dong, Zeqian Chen, Rui He, Chuan Chua, Caslon Chen, Feifei Yang, Yun Swinburne University of Technology Melbourne Australia Huazhong University of Science and Technology Wuhan China Deakin University Melbourne Australia National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster Grid Computing Lab School of Computer Science and Technology Huazhong University of Science and Technology Wuhan430074 China

ISBN: (纸本)9798400712746

In an edge-assisted federated learning (FL) system, edge servers aggregate the local models from the clients within their coverage areas to produce intermediate models for the production of the global model. This significantly reduces the communication overhead incurred during the FL process. To accelerate model convergence, FedEdge, the state-of-the-art edge-assisted FL system, trains clients’ models in local federations when they wait for the global model in each training round. However, our investigation reveals that it drives the global model towards clients with excessive local training, causing model drifts that undermine model performance for other clients. To tackle this problem, this paper presents Maverick, a new edge-assisted FL system that mitigates model drifts by training personalized local models for clients through contrastive local training. It introduces a model-contrastive loss to facilitate personalized local federated training by driving clients’ local models away from the global model and close to their corresponding intermediate models. In addition, Maverick includes anomalous models in contrastive local training as negative samples to accelerate the convergence of clients’ local models. Extensive experiments are conducted on three widely-used models trained on three datasets to comprehensively evaluate the performance of Maverick. Compared to state-of-the-art edge-assisted FL systems, Maverick accelerates model convergence by up to 16.2x and improves model accuracy by up to 12.7%. © 2025 Copyright held by the owner/author(s).

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

An Improved Detection of Skin Melanoma Using Hybrid CNN-MobileNet V2 Approach with Multi-Scale Attention

An Improved Detection of Skin Melanoma Using Hybrid CNN-Mobi...

引用

Machine Learning and Autonomous Systems (ICMLAS), International Conference on

作者： S.P. Maniraj R Thamizhamuthu Department of Data Science and Business Systems School of Computing SRM Institute of Science and Technology Chennai Tamilnadu India Department of Computing Technologies School of Computing SRM Institute of Science and Technology Chennai Tamilnadu India

ISBN: (数字)9798331505745

ISBN: (纸本)9798331505752

Identifying melanoma in dermoscopic images is an immense challenge due to the varied appearances of the lesions. This research introduces a hybrid CNN-LSTM model augmented with a multi-scale attention mechanism to tackle this complexity, utilising CNNs for spatial feature extraction and LSTMs for temporal pattern recognition. Our model surpasses traditional CNN and MobileNet V2-LSTM models, with an accuracy of 96.98%, a recall of 96.65%, a specificity of 97.56%, and a JSI of 95.78 % . These measurements demonstrate enhanced sensitivity and reliability compared to baseline models, signifying strong diagnostic capabilities. This study improves melanoma detection, aiding in more precise diagnoses for clinical environments and remote telemedicine applications. Future initiatives involve enhancing the model for more comprehensive skin lesion categorisation.

关键词： Accuracy Sensitivity Attention mechanisms Telemedicine Melanoma Sensitivity and specificity Skin Pattern recognition Lesions Reliability

来源：评论

学校读者我要写书评

暂无评论

Brain Tumor detection Using Machine Learning and Deep Learning Approaches

Brain Tumor detection Using Machine Learning and Deep Learni...

引用

2022 IEEE International Conference on Advances in computing, Communication and Applied Informatics, ACCAI 2022

作者： Khan, Oneza Tehreem Rajeswari, D. School of Computing College of Engineering and Technology Department of Data Science and Business Systems Kattankulathur603203 India

ISBN: (纸本)9781665495295

Recent years has provided extremely interesting and exciting developments and applications of Machine Learning (ML) and Deep Learning (DL) to automize healthcare delivery. ML techniques are surpassing human performances in detecting, predicting, diagnosing any disease and even recommending treatments for the same. ML along with Deep Learning has revolutionized Medical Imaging Technology. Brain Tumor detection is one of the many healthcare area Artificial Intelligence has marveled at. A tumor is basically an area or part of the body suffering by swelling caused by distorted tissues where the cells start multiplying uncontrollably. Magnetic Resonance Imaging (MRI) is the technique used to detect the tumor. MRI is quite commonly used in medical science to detect abnormalities in any part of the body. Tumor in the brain is far more perilous and different to treat than in any other part of the body which makes the early prediction and monitoring of brain tumor extremely expedient. This paper discusses the various algorithms, techniques, new approaches and their comparison with the existing ones and the general process of brain tumor detection with the help of MRI. © 2022 IEEE.

关键词： Magnetic resonance imaging

来源：评论

学校读者我要写书评

暂无评论

CVE-Bench: A Benchmark for AI Agents’ Ability to Exploit Real-World Web Application Vulnerabilities

arXiv

引用

arXiv 2025年

作者： Zhu, Yuxuan Kellermann, Antony Bowman, Dylan Li, Philip Gupta, Akul Danda, Adarsh Fang, Richard Jensen, Conner Ihli, Eric Benn, Jason Geronimo, Jet Dhir, Avi Rao, Sudhit Yu, Kaicheng Stone, Twm Kang, Daniel Siebel School of Computing and Data Science University of Illinois Urbana-Champaign United States

Large language model (LLM) agents are increasingly capable of autonomously conducting cyberattacks, posing significant threats to existing applications. This growing risk highlights the urgent need for a real-world benchmark to evaluate the ability of LLM agents to exploit web application vulnerabilities. However, existing benchmarks fall short as they are limited to abstracted Capture the Flag competitions or lack comprehensive coverage. Building a benchmark for real-world vulnerabilities involves both specialized expertise to reproduce exploits and a systematic approach to evaluating unpredictable threats. To address this challenge, we introduce CVE-Bench, a real-world cybersecurity benchmark based on critical-severity Common Vulnerabilities and Exposures. In CVE-Bench, we design a sandbox framework that enables LLM agents to exploit vulnerable web applications in scenarios that mimic real-world conditions, while also providing effective evaluation of their exploits. Our evaluation shows that the state-of-the-art agent framework can resolve up to 13% of vulnerabilities. Copyright © 2025, The Authors. All rights reserved.

关键词： Autonomous agents

来源：评论

学校读者我要写书评

暂无评论

To Invest or Not to Invest: Using Vocal Behavior to Predict Decisions of Investors in an Entrepreneurial Context 26th

To Invest or Not to Invest: Using Vocal Behavior to Predi...

引用

26th International Conference on Pattern Recognition, ICPR 2022

作者： Goossens, Ilona Jung, Merel M. Liebregts, Werner Onal Ertugrul, Itir Department of Cognitive Science and Artificial Intelligence Tilburg University Tilburg Netherlands Jheronimus Academy of Data Science ’s-Hertogenbosch Netherlands Department of Information and Computing Sciences Utrecht University Utrecht Netherlands

ISBN: (纸本)9783031376597

Entrepreneurial pitch competitions have become increasingly popular in the start-up culture to attract prospective investors. As the ultimate funding decision often follows from some form of social interaction, it is important to understand how the decision-making process of investors is influenced by behavioral cues. In this work, we examine whether vocal features are associated with the ultimate funding decision of investors by utilizing deep learning methods. We used videos of individuals in an entrepreneurial pitch competition as input to predict whether investors will invest in the startup or not. We proposed models that combine deep audio features and Handcrafted audio Features (HaF) and feed them into two types of Recurrent Neural Networks (RNN), namely Long Short-Term Memory (LSTM) and Gated Recurrent Units (GRU). We also trained the RNNs with only deep features to assess whether HaF provide additional information to the models. Our results show that it is promising to use vocal behavior of pitchers to predict whether investors will invest in their business idea. Different types of RNNs yielded similar performance, yet the addition of HaF improved the performance. © 2023, Springer Nature Switzerland AG.

关键词： Long short-term memory

来源：评论

学校读者我要写书评

暂无评论

CONTINUAL LEARNING WITH EMBEDDING LAYER SURGERY AND TASK-WISE BEAM SEARCH USING WHISPER

arXiv

引用

arXiv 2025年

作者： Kwok, Chin Yuen Yip, Jia Qi Chng, Eng Siong College of Computing and Data Science Nanyang Technological University Singapore Alibaba Group Singapore

Current Multilingual ASR models only support a fraction of the world’s languages. Continual Learning (CL) aims to tackle this problem by adding new languages to pre-trained models while avoiding the loss of performance on existing languages, also known as Catastrophic Forgetting (CF). However, existing CL methods overlook the adaptation of the token embedding lookup table at the decoder, despite its significant contribution to CF. We propose Embedding Layer Surgery where separate copies of the token embeddings are created for each new languages, and one of the copies is selected to replace the old languages embeddings when transcribing the corresponding new language. Unfortunately, this approach means LID errors also cause incorrect ASR embedding selection. Our Task-wise Beam Search allows self-correction for such mistakes. By adapting Whisper to 10 hours of data for each of 10 unseen languages from Common Voice, results show that our method reduces the Average WER (AWER) of pre-trained languages from 14.2% to 11.9% compared with Experience Replay, without compromising the AWER of the unseen languages. © 2025, CC BY-NC-ND.

关键词： Table lookup

来源：评论

学校读者我要写书评

暂无评论

Do Body Expressions Leave Good Impressions? - Predicting Investment Decisions based on Pitcher's Body Expressions 25

Do Body Expressions Leave Good Impressions? - Predicting Inv...

引用

25th International Conference on Multimodal Interaction, ICMI 2023 Companion

作者： Jung, Merel M. Van Vlierden, Mark Liebregts, Werner Önal Ertuǧrul, Itlr Department of Cognitive Science and Artifcial Intelligence Tilburg University Tilburg Netherlands Jheronimus Academy of Data Science s-Hertogenbosch Netherlands Department of Information and Computing Sciences Utrecht University Utrecht Netherlands

ISBN: (纸本)9798400703218

This work aims to investigate whether displayed body expressions during venture pitches hold predictive value over the success of gathering funding. To this end the performance of several traditional regression models trained on aggregated body expression features is compared to that of deep recurrent models. Moreover, the impact of features and the effect of oversampling to counteract data imbalance are investigated. The results showed that temporal modeling of body expressions with GRU yielded the best performance (average MAE = 16.9). Oversampling was found to not improve models' performance. The analysis of the impact of features suggest that higher levels of expressed body movements negatively affect investment scores whereas a more open body posture can have a positive effect. These findings show that body expressions can be predictive of investment decisions. Future work will be directed towards integrating these cues into a multimodal model to aid entrepreneurs in improving their pitching skills and further the understanding of the investment decision-making process. © 2023 Owner/Author.

关键词： Investments

来源：评论

学校读者我要写书评

暂无评论

data-Adaptive Low-Rank Sparse Subspace Clustering

arXiv

引用

arXiv 2025年

作者： Kopriva, Ivica Division of Computing and Data Science Ruđer Bošković Institute Zagreb Croatia

Low-rank sparse subspace clustering (LRSSC) algorithms built on self-expressive model effectively capture both the global and local structure of the data. However, existing solutions, primarily based on proximal operators associated with Sp / l p , p∈{0, 1/2, 2/3, 1}, norms are not data-adaptive. In this work, we propose an LRSSC algorithm incorporating a data-adaptive surrogate for the S0 / l 0 quasi-norm. We provide a numerical solution for the corresponding proximal operator in cases where an analytical expression is unavailable. The proposed LRSSC algorithm is formulated within the proximal mapping framework, and we present theoretical proof of its global convergence toward a stationary point. We evaluate the performance of the proposed method on three well known datasets, comparing it against LRSSC algorithms constrained by Sp / l p , p∈{0, 1/2, 2/3, 1}, *** Codes 68W05, 68U10, 65Y99 © 2025, CC BY.

关键词： data assimilation

来源：评论

学校读者我要写书评

暂无评论

IoT Smart City Air Quality Monitoring System Using Arduino

IoT Smart City Air Quality Monitoring System Using Arduino

引用

2023 Intelligent computing and Control for Engineering and Business Systems, ICCEBS 2023

作者： Gopinath, N. Chinnasamy, A. Sachin, K. Sakthiraj, S. Ravindharan, S. Sri Sairam Engineering College Department of Computer Science and Engineering Tamil Nadu Chennai India School of Computing Srmist Kattankulathur Campus Department of Data Science and Business Systems Chennai India

ISBN: (纸本)9798350394580

Many factors, including population growth, increased vehicle use, industrialization, and urbanisation, have contributed to an increase in pollution levels throughout time, which has a negative impact on human wellbeing by adversely affecting the health of those exposed to it. To keep an eye on In this project, we'll build an IOT-based air pollution monitoring system that will track the air quality via an internet-connected web server and sound an alarm whenit drops below a certain threshold, which occurs when harmful gases like CO2, smoke, alcohol, benzene, and NH3 are present in sufficient quantities. On the LCD and on the website, the air quality will be displayed in PPM for easy monitoring. You can use your mobiledevice to check the pollution level as part of this IOT project from anywhere. © 2023 IEEE.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Traffic Sign Detection Using YOLOv8

Traffic Sign Detection Using YOLOv8

引用

Intelligent data Communication Technologies and Internet of Things (IDCIoT), International Conference on

作者： Rahul Kumar Aniket Gupta Rajeswari D Department of Data Science and Business Systems School of computing College of Engineering and Technology

Traffic signals and other signs like parking., stop signs., etc. have become very crucial in autonomous and s elf-driving cars as it helps the smart system to comply with the basic traffic rules along with that it helps navigate routes based on the signs thus enabling a more secure driving experience for the drivers. There have been a lot of new algorithms that have emerged in the past recent years regarding this. In this paper., this research has used the new YOLOv8 object detection system to help us detect traffic signs as it is much fas ter and more precis e than its previous iterations. To improve the algorithm., this paper has used a dataset comprising photos of traffic signs taken at different angles and different light intensities. This system can predict the traffic signs with 93% accuracy.

关键词：

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：