Research and Application of the Median Filtering Method in Enhancing the Imperceptibility of Perturbations in Adversarial Examples DOI Open Access
Yiming He, Yanhua Dong, Hongyu Sun

et al.

Electronics, Journal Year: 2024, Volume and Issue: 13(13), P. 2458 - 2458

Published: June 23, 2024

In the field of object detection, adversarial attack method based on generative network efficiently generates examples, thereby significantly reducing time costs. However, this approach overlooks imperceptibility perturbations in resulting poor visual performance and insufficient invisibility generated examples. To further enhance a utilizing median filtering is proposed to address these perturbations. Experimental evaluations were conducted Pascal VOC dataset. The results demonstrate that, compared original image, there an increase at least 17.2% structural similarity index (SSIM) for Additionally, peak signal-to-noise ratio (PSNR) increases by 27.5%, while learned perceptual image patch (LPIPS) decreases 84.6%. These findings indicate that examples are more difficult detect, with improved closer resemblance without compromising their high aggressiveness.

Language: Английский

ChatGPT: A Case Study on Copyright Challenges for Generative Artificial Intelligence Systems DOI Creative Commons
Nicola Lucchi

European Journal of Risk Regulation, Journal Year: 2023, Volume and Issue: 15(3), P. 602 - 624

Published: Aug. 29, 2023

Abstract This article focuses on copyright issues pertaining to generative artificial intelligence (AI) systems, with particular emphasis the ChatGPT case study as a primary exemplar. In order generate high-quality outcomes, AI systems require substantial quantities of training data, which may frequently comprise copyright-protected information. prompts inquiries into legal principles fair use, creation derivative works and lawfulness data gathering utilisation. The utilisation input for purpose enhancing models presents significant concerns regarding potential violations copyright. paper offers suggestions safeguarding interests holders competitors, while simultaneously addressing challenges expediting advancement technologies. analyses platform example explore necessary modifications that regulations must undergo adequately tackle intricacies authorship ownership in realm AI-generated creative content.

Language: Английский

Citations

63
Next gen cybersecurity paradigm towards artificial general intelligence: Russian market challenges and future global technological trends DOI
Е. В. Плешакова, Aleksey Osipov, Sergey Gataullin

et al.

Journal of Computer Virology and Hacking Techniques, Journal Year: 2024, Volume and Issue: 20(3), P. 429 - 440

Published: July 23, 2024

Language: Английский

Citations

31
A pilot study of measuring emotional response and perception of LLM-generated questionnaire and human-generated questionnaires DOI Creative Commons
Zhao Zou, Omar Mubin, Fady Alnajjar

et al.

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: Feb. 2, 2024

Abstract The advent of ChatGPT has sparked a heated debate surrounding natural language processing technology and AI-powered chatbots, leading to extensive research applications across various disciplines. This pilot study aims investigate the impact on users' experiences by administering two distinct questionnaires, one generated humans other ChatGPT, along with an Emotion Detecting Model. A total 14 participants (7 female 7 male) aged between 18 35 years were recruited, resulting in collection 8672 ChatGPT-associated data points 8797 human-associated points. Data analysis was conducted using Analysis Variance (ANOVA). results indicate that utilization enhances participants' happiness levels reduces their sadness levels. While no significant gender influences observed, variations found about specific emotions. It is important note limited sample size, narrow age range, potential cultural impacts restrict generalizability findings broader population. Future directions should explore incorporating additional models or chatbots user emotions, particularly among groups such as older individuals teenagers. As pioneering works evaluating human perception text communication, it noteworthy received positive evaluations demonstrated effectiveness generating questionnaires.

Language: Английский

Citations

18
AI in Breast Cancer Imaging: An Update and Future Trends DOI Creative Commons
Yizhou Chen, Xiaoliang Shao, Kuangyu Shi

et al.

Seminars in Nuclear Medicine, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 1, 2025

Language: Английский

Citations

3
Error Correction and Adaptation in Conversational AI: A Review of Techniques and Applications in Chatbots DOI Creative Commons

Saadat Izadi,

Mohamad Forouzanfar

AI, Journal Year: 2024, Volume and Issue: 5(2), P. 803 - 841

Published: June 4, 2024

This study explores the progress of chatbot technology, focusing on aspect error correction to enhance these smart conversational tools. Chatbots, powered by artificial intelligence (AI), are increasingly prevalent across industries such as customer service, healthcare, e-commerce, and education. Despite their use increasing complexity, chatbots prone errors like misunderstandings, inappropriate responses, factual inaccuracies. These issues can have an impact user satisfaction trust. research provides overview chatbots, conducts analysis they encounter, examines different approaches rectifying errors. include using data-driven feedback loops, involving humans in learning process, adjusting through methods reinforcement learning, supervised unsupervised semi-supervised meta-learning. Through real life examples case studies fields, we explore how strategies implemented. Looking ahead, challenges faced AI-powered including ethical considerations biases during implementation. Furthermore, transformative potential new technological advancements, explainable AI models, autonomous content generation algorithms (e.g., generative adversarial networks), quantum computing training. Our information for developers researchers looking improve capabilities, which be applied service support effectively address requirements.

Language: Английский

Citations

15
A Comprehensive Review of Deep Learning: Architectures, Recent Advances, and Applications DOI Creative Commons
Ibomoiye Domor Mienye, Theo G. Swart

Information, Journal Year: 2024, Volume and Issue: 15(12), P. 755 - 755

Published: Nov. 27, 2024

Deep learning (DL) has become a core component of modern artificial intelligence (AI), driving significant advancements across diverse fields by facilitating the analysis complex systems, from protein folding in biology to molecular discovery chemistry and particle interactions physics. However, field deep is constantly evolving, with recent innovations both architectures applications. Therefore, this paper provides comprehensive review DL advances, covering evolution applications foundational models like convolutional neural networks (CNNs) Recurrent Neural Networks (RNNs), as well such transformers, generative adversarial (GANs), capsule networks, graph (GNNs). Additionally, discusses novel training techniques, including self-supervised learning, federated reinforcement which further enhance capabilities models. By synthesizing developments identifying current challenges, insights into state art future directions research, offering valuable guidance for researchers industry experts.

Language: Английский

Citations

15
Securing tomorrow: a comprehensive survey on the synergy of Artificial Intelligence and information security DOI Creative Commons
Ehtesham Hashmi, Muhammad Mudassar Yamin, Sule Yildirim Yayilgan

et al.

AI and Ethics, Journal Year: 2024, Volume and Issue: unknown

Published: July 30, 2024

Abstract This survey paper explores the transformative role of Artificial Intelligence (AI) in information security. Traditional methods, especially rule-based approaches, faced significant challenges protecting sensitive data from ever-changing cyber threats, particularly with rapid increase volume. study thoroughly evaluates AI’s application security, discussing its strengths and weaknesses. It provides a detailed review impact on examining various AI algorithms used this field, such as supervised, unsupervised, reinforcement learning, highlighting their respective limitations. The identifies key areas for future research focusing improving algorithms, strengthening addressing ethical issues, exploring safety security-related concerns. emphasizes security risks, including vulnerability to adversarial attacks, aims enhance robustness reliability systems by proposing solutions potential threats. findings aim benefit cybersecurity professionals researchers offering insights into intricate relationship between AI, emerging technologies.

Language: Английский

Citations

12
Comprehensive review and comparative analysis of transformer models in sentiment analysis DOI
Hadis Bashiri, Hassan Naderi

Knowledge and Information Systems, Journal Year: 2024, Volume and Issue: unknown

Published: Sept. 6, 2024

Language: Английский

Citations

9
EXPRESS: The Role of National Culture: an Updated Framework for Cross-Cultural Research in Operations Management DOI
Manjul Gupta, Sushil Gupta

Production and Operations Management, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 31, 2025

National culture plays a vital role in shaping operations management (OM) practices, yet its impact remains largely underexplored. This paper builds on the Gupta and Gupta’s cross-cultural OM research framework that consists of three categories: operational decisions, supply chain management, interdisciplinary topics. It highlights how national influences key areas such as product return policies, dynamic complexity, buyer–supplier conflict relationships. Additionally, it examines culture's determining post-acquisition performance cross-border mergers acquisitions, adoption digital piracy prevention strategies, relationship between language chains. To address emerging challenges, extends by introducing new themes. concludes with recommendations for future research, offering valuable guidance scholars practitioners navigating complexities managing culturally diverse globally interconnected operations.

Language: Английский

Citations

1
Learning and Teaching in the Era of Generative Artificial Intelligence Technologies: An In‐Depth Exploration Using Multi‐Analytical SEM‐ANN Approach DOI Open Access
Muhammad Farrukh Shahzad, Shuo Xu, Xin An

et al.

European Journal of Education, Journal Year: 2025, Volume and Issue: 60(1)

Published: Feb. 18, 2025

ABSTRACT The arrival of generative artificial intelligence (GAI) technologies marks a significant transformation in the educational landscape, with implications for teaching and learning performance. These can generate content, simulate interactions, adapt to learners' needs, offering opportunities interactive experiences. In China's education sector, incorporating GAI address challenges, enhance practices, improve This study scrutinises impact on performance focusing mediating roles e‐learning competence (EC), desire (DL), beliefs about future (BF), as well moderating role facilitating conditions amongst Chinese educators. Data was collected from 411 teachers across various institutions China using purposive sampling. PLS‐SEM ANN were employed assess suggested structural model. results indicate that significantly influence by EC, DL, BF roles. Furthermore, positively moderate association BF. underscores critical self‐determination theory shaping effective incorporation education, valuable insights outcomes sector.

Language: Английский

Citations

1