检索结果-内蒙古大学图书馆

IEEE Transactions on Technology and Society 2025年第1期6卷 64-79页

作者： Twomey, John Ching, Didier Aylett, Matthew Peter Quayle, Michael Linehan, Conor Murphy, Gillian University College Cork School of Applied Psychology CorkT12 CY82 Ireland Lero—The Science Foundation Ireland Research Centre for Software LimerickV94 NYD3 Ireland Heriot Watt University Mathematics and Computer Science Department EdinburghEH14 4AS United Kingdom CereProc Ltd. EdinburghEH3 9DR United Kingdom University of Limerick Centre for Social Issues Research Department of Psychology LimerickV94 T9PX Ireland University of KwaZulu-Natal School of Applied Human Sciences Department of Psychology Pietermaritzburg3201 South Africa

Deepfakes are a form of synthetic media that uses deep-learning technology to create fake images, video, and audio. The emergence of this technology has inspired much commentary and speculation from academics across a range of disciplines, who have contributed expert opinions regarding the implications of deepfake proliferation on fields such as law, politics, and entertainment. A systematic scoping review was carried out to identify, assemble, and critically analyze those academic narratives. The aim is to build on and critique previous attempts at defining the technology and categorizing the harms and benefits of deepfake technology. A range of databases were searched for relevant articles from 2017 to 2023, resulting in a large multi-disciplinary dataset of 102 papers, 181,659 words long, which were analyzed qualitatively through thematic analysis. Implications for future research include questioning the lack of research evidence for the supposed positives of deepfakes, recognizing the role that identity plays in deepfake technology, challenging the perceived accessibility/ believability of deepfakes, and proposing a more nuanced approach to the dichotomous "positive and negatives" of deepfakes. Furthermore, we show how definitional issues around what a deepfake is versus other forms of fake media feeds confusion around the novelty and impacts of deepfakes. © 2024 The Authors.

关键词： Deepfakes Reviews Generative AI Software Psychology Social networking (online) Organizations Lenses Law Internet

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

Synthesising knocking sound effects using Conditional WaveGAN 17

Synthesising knocking sound effects using Conditional WaveGA...

引用

17th Sound and Music Computing Conference, SMC 2020

作者： Barahona-Róos, Adrián Pauletto, Sandra Department of Computer Science University of York Department of Media Technology and Interaction Design KTH Royal Institute of Technology

ISBN: (纸本)9788894541502

In this paper we explore the synthesis of sound effects using conditional generative adversarial networks (cGANs). We commissioned Foley artist Ulf Olausson to record a dataset of knocking sound effects with different emotions and trained a cGAN on it. We analysed the resulting synthesised sound effects by comparing their temporal acoustic features to the original dataset and by performing an online listening test. Results show that the acoustic features of the synthesised sounds are similar to those of the recorded dataset. Additionally, the listening test results show that the synthesised sounds can be identified by people with experience in sound design, but the model is not far from fooling non-experts. Moreover, on average most emotions can be recognised correctly in both recorded and synthesised sounds. Given that the temporal acoustic features of the two datasets are highly similar, we hypothesise that they strongly contribute to the perception of the intended emotions in the recorded and synthesised knocking sounds. Copyright © 2020 Adrián Barahona-Ríos et al.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Summon a Demon and Bind It: A Grounded Theory of LLM Red Teaming

arXiv

引用

arXiv 2023年

作者： Inie, Nanna Stray, Jonathan Derczynski, Leon Paul G. Allen School of Computer Science & Engineering SeattleWA United States Department of Computer Science IT University of Copenhagen Copenhagen Denmark Berkeley Center for Human-Compatible AI UC Berkeley BerkeleyCA United States NVIDIA Corporation United States

Engaging in the deliberate generation of abnormal outputs from Large Language Models (LLMs) by attacking them is a novel human activity. This paper presents a thorough exposition of how and why people perform such attacks, defining LLM red-teaming based on extensive and diverse evidence. Using a formal qualitative methodology, we interviewed dozens of practitioners from a broad range of backgrounds, all contributors to this novel work of attempting to cause LLMs to fail. We focused on the research questions of defining LLM red teaming, uncovering the motivations and goals for performing the activity, and characterizing the strategies people use when attacking LLMs. Based on the data, LLM red teaming is defined as a limit-seeking, non-malicious, manual activity, which depends highly on a team-effort and an alchemist mindset. It is highly intrinsically motivated by curiosity, fun, and to some degrees by concerns for various harms of deploying LLMs. We identify a taxonomy of 12 strategies and 35 different techniques of attacking LLMs. These findings are presented as a comprehensive grounded theory of how and why people attack large language models: LLM red teaming. © 2023, CC BY.

关键词： Computational linguistics

来源：评论

学校读者我要写书评

暂无评论

Collaborative Synergy: Enhancing Face-to-Face Computing Courses

Collaborative Synergy: Enhancing Face-to-Face Computing Cour...

引用

Frontiers in Education (FIE) Conference

作者： Iryna Ashby Deepti Tagare Cassandra Thomas Marisa Exter College of Education Purdue University West Lafayette USA College of Education and Human Development University of Texas at San Antonio San Antonio USA Department of Computer Science Tuskegee University Tuskegee USA

ISBN: (数字)9798350351507

ISBN: (纸本)9798350363067

This innovative practice paper explores the varied perspectives of a computing faculty member and a group of instructional designers, who partnered to revise courses to increase active learning practices and integrate cross-disciplinary skills and dispositions into two face-to-face computing courses (CS 1 and CS 2), as part of a larger, grant-funded systemic change effort. The instructional design team members varied in terms of level of formal educational experience in engineering, computing, and instructional design

关键词： Education Active learning Collaboration Engineering computing

来源：评论

学校读者我要写书评

暂无评论

Federated Learning and Convolutional Neural Networks for Intrusion Detection Systems

Federated Learning and Convolutional Neural Networks for Int...

引用

International Conference on Pattern Analysis and Intelligent Systems (PAIS)

作者： Malika Abid Mohammed Kamel Benkaddour Mohamed Benouis Computer Science Department Faculty of Sciences and Technology Artificial Intelligence and Information Technology Laboratory (LINATI) University of Kasdi Merbah Ouargla Algeria Human-Centered Artificial Intelligence University of Augsburg Augsburg Germany

ISBN: (数字)9798350350265

ISBN: (纸本)9798350350272

Recently, there has been a growing focus on network traffic threats due to the proliferation of mobile devices and the expanding Internet of Things (IoT). The centralised anomaly detection approach implies sharing the user’s data with a cloud server, which mostly exposes the user to the data breach issue. Given the above issues, this paper proposes a federated learning intrusion detection system (FedIDS) to improve the performance of intrusion detection while protecting user privacy. Motivated by recent exploits into deep learning for intrusion detection, we propose a convolutional neural network (CNN) model to learn and train the record data collected from independent and non-independent identically distributed (Non-IID, IID)data. This work conducts experiments on the NSL-KDD Data set for network intrusion detection. Experimental results show that FedIDS has achieved a competitive advantage over state-of-the-art centralised machine learning methods.

关键词： Performance evaluation Analytical models Privacy Federated learning Intrusion detection Telecommunication traffic Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

FoamFactor: Hydrogel-foam composite with tunable stiffness and compressibility

arXiv

引用

arXiv 2021年

作者： Yang, Humphrey Yan, Zeyu Luo, Danli Yao, Lining Human-Computer Interaction Carnegie Mellon University PittsburghPA United States Computer Science University of Maryland College Park MD United States

This paper presents FoamFactor, a novel material with tunable stiffness and compressibility between hydration states, and a tailored pipeline to design and fabricate artifacts consisting of it. This technique compounds hydrogel with open-cell foams via additive manufacturing to produce a water-responsive composite material. Enabled by the large volumetric changes of hydrogel dispersions, the material is soft and compressible when dehydrated and becomes stiffer and rather incompressible when hydrated. Leveraging this material property transition, we explore its design space in various aspects pertaining the transition of hydration states, including multi-functional shoes, amphibious cars, mechanical transmission systems, and self-deploying robotic grippers. © 2021, CC BY.

关键词： Hydrogels

来源：评论

学校读者我要写书评

暂无评论

Domain Knowledge-Driven Multi-Label Behavioral Health Identification from Police Report

Domain Knowledge-Driven Multi-Label Behavioral Health Identi...

引用

IEEE International Conference on Big Data

作者： Abm Adnan Azmee Francis Nweke Md Abdullah Al Hafiz Khan Yong Pei Dominic Thomas Monica Nandan Department of Computer Science Kennesaw State University Marietta Georgia USA Department of Information Systems and Security Kennesaw State University Kennesaw Georgia USA Department of Social Work and Human Services Kennesaw State University Kennesaw Georgia USA

ISBN: (数字)9798350362480

ISBN: (纸本)9798350362497

Behavioral health plays a pivotal role in individuals’ overall quality of life. Timely identification and intervention of behavioral health concerns are essential for building a supportive community. In case of emergencies, first responders (police, fire, EMT) provide support to the community. The report of the first responders contains valuable cues about individuals struggling with behavioral health concerns. Currently, first responders manually screen these reports to determine individuals with behavioral health concerns. Manual screening of these reports is time-consuming and prone to error. Automated analysis of these reports is essential for providing timely intervention. Natural language processing techniques demonstrated potential for automated analysis of textual data. However, due to the unstructured and complex nature of the data, it’s challenging for traditional deep learning and natural language processing techniques to identify these cases. Moreover, support from subject matter experts is essential for the effective identification of these cases. In this research, we have developed a multi-label behavioral health detection framework that utilizes domain knowledge from subject matter experts for the effective identification of behavioral health cases. In addition, we have expanded the domain knowledge base by recognizing new behavioral health-related terms. We validated the efficacy and effectiveness of our proposed model through a comprehensive evaluation of real-world data. Furthermore, our proposed model demonstrated 6-24% performance improvement over the state-of-the-art models.

关键词： Deep learning Law enforcement Subject matter experts Knowledge based systems Collaboration Multi label classification Manuals Transformers Natural language processing Data models

来源：评论

学校读者我要写书评

暂无评论

Training Variational Autoencoders for Population Synthesis in Public Health with Missing Data

Training Variational Autoencoders for Population Synthesis i...

引用

IEEE International Conference on Big Data

作者： Nathan Blackthorn Andrew Arash Mahyari Ashok Srinivasan Department of Computer Science University of West Florida Pensacola FL USA Florida Institute For Human and Machine Cognition (IHMC) Pensacola FL USA Department of Intelligent Systems and Robotics University of West Florida Pensacola FL USA

ISBN: (数字)9798350362480

ISBN: (纸本)9798350362497

The importance of social determinants of health in shaping equitable public health policies is gaining increasing recognition. Emerging data sources, such as mobility and social media data, are becoming key to public health models. However, privacy concerns often limit access to these sensitive data, as even anonymized datasets are vulnerable to deductive disclosure. A promising solution to this challenge is the use of synthetic populations, which not only safeguard privacy but also enable the exploration of various "what-if" scenarios. Most existing studies on synthetic populations assume full access to complete datasets for training. However, in many public health applications, such as census data, this assumption is unrealistic due to incomplete responses, especially regarding sensitive questions like wealth or education. This paper introduces a novel approach for training Variational AutoEncoders (VAEs) with incomplete data, without resorting to missing value imputation. Instead, the VAEs are trained solely on the observed data. Using the 2019 PUMS dataset for Florida, we successfully train VAEs to generate diverse and flexible synthetic populations. By comparing marginal distributions and utilizing t-SNE for analysis, the results highlight the effectiveness of this method in addressing missing data challenges. This work demonstrates the potential of VAEs in generating synthetic populations for health research, even when complete datasets are unavailable, thereby offering a robust solution to advance public health studies while preserving data privacy.

关键词： Training Data privacy Privacy Social networking (online) Soft sensors Autoencoders Refining Imputation Data models Public healthcare

来源：评论

学校读者我要写书评

暂无评论

Searching thousands of genomes to classify somatic and novel structural variants using STIX (vol 19, pg 445, 2022)

引用

NATURE METHODS 2022年第6期19卷 770-770页

作者： Chowdhury, Murad Pedersen, Brent S. Sedlazeck, Fritz J. Quinlan, Aaron R. Layer, Ryan M. BioFrontiers Institute University of Colorado Boulder CO USA University Medical Center Utrecht University Utrecht the Netherlands Human Genome Sequencing Center Baylor College of Medicine Houston TX USA Department of Human Genetics University of Utah Salt Lake City UT USA Department of Biomedical Informatics University of Utah Salt Lake City UT USA Utah Center for Genetic Discovery University of Utah Salt Lake City UT USA Department of Computer Science University of Colorado Boulder CO USA

Structural variants are associated with cancers and developmental disorders, but challenges with estimating population frequency remain a barrier to prioritizing mutations over inherited variants. In particular, variability in variant calling heuristics and filtering limits the use of current structural variant catalogs. We present STIX, a method that, instead of relying on variant calls, indexes and searches the raw alignments from thousands of samples to enable more comprehensive allele frequency estimation.

关键词： Genome informatics Genomics Mutation Software Life sciences general Biological Techniques Biological Microscopy Biomedical Engineering/Biotechnology Bioinformatics Proteomics

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：