检索结果-内蒙古大学图书馆

17th Asian Conference on computer Vision, ACCV 2024

作者： Kim, Jonghyeon Ju, Chan-Yang Kim, Gun-Woo Lee, D.-H. Department of Applied Artificial Intelligence Hanyang University Ansan Korea Republic of Department of Computer Science and Engineering Gyeongsang National University Jinju Korea Republic of

ISBN: (纸本)9789819609000

Multi-face tracking (MFT) is a subtask of multi-object tracking (MOT) that focuses on detecting and tracking multiple faces across video frames. Modern MOT trackers adopt the Kalman filter (KF), a linear model that estimates current motions based on previous observations. However, these KF-based trackers struggle to predict motions in unconstrained videos with frequent shot changes, occlusions, and appearance variations. To address these limitations, we propose BoT-FaceSORT, a novel MFT framework that integrates shot change detection, shared feature memory, and an adaptive cascade matching strategy for robust tracking. It detects shot changes by comparing the color histograms of adjacent frames and resets KF states to handle discontinuities. Additionally, we introduce MovieShot, a new benchmark of challenging movie clips to evaluate MFT performance in unconstrained scenarios. We also demonstrate the superior performance of our method compared to existing methods on three benchmarks, while an ablation study validates the effectiveness of each component in handling unconstrained videos. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Kalman filters

来源：评论

学校读者我要写书评

暂无评论

Revisiting Location Privacy in MEC-Enabled Computation Offloading

引用

IEEE Transactions on Information Forensics and Security 2025年 20卷 4396-4407页

作者： Li, Jingyi Ou, Wenzhong Ouyang, Bei Ye, Shengyuan Zeng, Liekang Chen, Lin Chen, Xu Sun Yat-sen University School of Computer Science and Engineering Guangzhou510275 China Chinese University of Hong Kong Department of Information Engineering Hong Kong Macao Polytechnic University Engineering Research Centre of Applied Technology on Machine Translation and Artificial Intelligence China

Mobile Edge Computing (MEC) revolutionizes real-time applications by extending cloud capabilities to network edges, enabling efficient computation offloading from mobile devices. In recent years, the location privacy concern within MEC offloading has been recognized, prompting the proposal of various methodologies to mitigate this concern. However, this paper demonstrates that the prevailing privacy protection methods exhibit vulnerabilities. First, we analyze the shortcomings of current methodologies through both system modeling and evaluation metrics. Then, we introduce a Learning-based Trajectory Reconstruction Attack (LTRA) to expose the weaknesses, achieving up to 91.2% reconstruction accuracy against the state-of-the-art protection method. Further, based on w-event differential privacy, we propose an l-trajectory differentially private mechanism, i.e., OffloadingBD. Compared to the existing works, OffloadingBD provides more flexible and enhanced protection with sound privacy theoretical guarantee. Lastly, we conduct extensive experiments to evaluate LTRA and OffloadingBD. The experiment results show that LTRA has good generalization ability and OffloadingBD showcases a superior balance between privacy and utility compared with baselines. © 2005-2012 IEEE.

关键词： Differential privacy

来源：评论

学校读者我要写书评

暂无评论

To Guide or to Disturb - How to Teach Dexterous Skills Using AI? 25

To Guide or to Disturb - How to Teach Dexterous Skills Using...

引用

30th International Conference on Intelligent User Interfaces, IUI 2025

作者： Park, Jiyoung Kwak, Dohoon Lim, Jaewan Lee, Taeyoon Kim, Seung-Gyeom Yoo, Yongjae Department of Applied Artificial Intelligence Hanyang University Gyeonggi-do Ansan-si Korea Republic of Department of Artificial Intelligence Hanyang University Gyeonggi-do Ansan-si Korea Republic of Division of Computer Science Hanyang University Gyeonggi-do Ansan-si Korea Republic of

ISBN: (纸本)9798400713064

Learning a skill is a complex cognitive task that requires processing multiple sensorimotor information. To help the skill learners, in this study, we explored and compared how to exploit AI wisely to teach skills. We set an initial target skill of engraving implemented on a haptic-audio-visual (HAV) VR environment. As the haptic device, we used a desktop force-feedback device to render the movement and force profile of the *** build teaching AI for engraving, we gathered experts' motion and force profile data. Then, we designed an Long-Short Term Memory (LSTM)-based AI model that discriminates the user's behavior and status of the tooltip using the data. With the VR environment and the AI model, we compared and evaluated three teaching strategies for haptic dexterous skill transfer- Guidance, Disturbance, and Hybrid Assistance. Hybrid Assistance alters its force between Guidance and Disturbance based on the user's performance. We conducted a user experiment with seven sessions: one pre-test, three main training, one immediate retention, and two delayed retention sessions. In the results, we found: 1) Guidance showed a steep learning curve during the training sessions, but the participants lost the learning effect in the retention sessions, and 2) the learning with Hybrid Assistance was the slowest but remained longer, even showed a better performance in delayed retention tests. These results seem to follow a guidance hypothesis in learning, which suggests how to design the AI model to determine its policy to provide the best performance for the user's training. © 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

Vulnerability Analysis of IoT Platforms: A Case Study on *** 24

Vulnerability Analysis of IoT Platforms: A Case Study on ***

引用

14th International Conference on the Internet of Things, IoT 2024

作者： Molina, Jose M. Bustamante, Alvaro Patricio, Miguel Angel Computer Science and Engineering Department Applied Artificial Intelligence Group Universidad Carlos III de Madrid Colmenarejo Madrid Spain Thinger.io Engineering Team Las Rozas Madrid Spain

ISBN: (纸本)9798400712852

The rapid growth of the Internet of Things (IoT) has introduced significant security challenges due to the variety of communication protocols and the often-limited security measures in many devices. This paper presents a comprehensive set of techniques for evaluating the security of IoT protocols, including guided testing, fuzzing, symbolic execution, formal verification, and penetration testing. As a practical demonstration, guided testing and fuzzing are applied to the open-source *** platform to assess its communication protocols, uncovering critical vulnerabilities in WebSockets, MQTT, and proprietary protocols. The results illustrate the effectiveness of these methods in identifying security flaws and offer recommendations to improve the resilience of IoT systems. This work contributes to ongoing efforts to secure IoT ecosystems and cyber-physical systems by providing actionable insights and a methodology that can be applied across different platforms. © 2024 Copyright held by the owner/author(s).

关键词： Formal verification

来源：评论

学校读者我要写书评

暂无评论

How Large Language Models are Transforming Teachers' Assessment of Student Competency: A Case Study on LLM-Based Report Writing

How Large Language Models are Transforming Teachers' Assessm...

引用

2025 International Conference on Electronics, Information, and Communication, ICEIC 2025

作者： Lee, Dongyub Kim, Daejung Loeser, Martin Seo, Kyoungwon Seoul National University of Science and Technology Department of Applied Artificial Intelligence Seoul Korea Republic of Seoul National University of Science and Technology Department of Environmental Engineering Seoul Korea Republic of Electrical Engineering and Mechatronics ZHAW Zurich University of Applied Sciences Department of Computer Science Winterthur Switzerland

ISBN: (纸本)9798331510756

Evaluating student competencies is critical to fostering academic success, but the traditional process of assessment and report generation is time-consuming and labor-intensive for teachers. The emergence of Large Language Models (LLMs) presents an opportunity to automate this process, potentially reducing the workload for educators. However, concerns such as hallucinations in LLM outputs and the lack of usability studies involving teachers raise questions about their reliability and practical application. In response, we developed an LLM-based report writing system specifically designed for real-time competency evaluation. To assess its effectiveness, we conducted a case study involving five educational experts, each with over 15 years of experience. These experts used the system to generate student competency reports and found them to be both sensible and specific. While they expressed concerns about issues like insufficient detail, they recognized the system's significant time-saving benefits. Our findings demonstrate that LLMs can positively impact teachers' competency assessments by streamlining the reporting process, offering valuable support for educators. © 2025 IEEE.

关键词： Students

来源：评论

学校读者我要写书评

暂无评论

PromptMap: An Alternative Interaction Style for AI-Based Image Generation 25

PromptMap: An Alternative Interaction Style for AI-Based Ima...

引用

30th International Conference on Intelligent User Interfaces, IUI 2025

作者： Adamkiewicz, Krzysztof Woźniak, Pawel Wojciech Dominiak, Julia Romanowski, Andrzej Karolus, Jakob Frolov, Stanislav Institute of Applied Computer Science Lodz University of Technology Lodz Poland Tu Wien Vienna Austria Lodz University of Technology Lodz Poland German Research Center for Artificial Intelligence Kaiserslautern Germany

ISBN: (纸本)9798400713064

Recent technological advances popularized the use of image generation among the general public. Crafting effective prompts can, however, be difficult for novice users. To tackle this challenge, we developed PromptMap, a new interaction style for text-to-image AI that allows users to freely explore a vast collection of synthetic prompts through a map-like view with semantic zoom. PromptMap groups images visually by their semantic similarity, allowing users to discover relevant examples. We evaluated PromptMap in a between-subject online study (n = 60) and a qualitative within-subject study (n = 12). We found that PromptMap supported users in crafting prompts by providing them with examples. We also demonstrated the feasibility of using LLMs to create vast example collections. Our work contributes a new interaction style that supports users unfamiliar with prompting in achieving a satisfactory image output. © 2025 Copyright held by the owner/author(s).

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Federated Knowledge Distillation With 3D Transformer Adaptation for Weakly Labeled Multi-Organ Medical Image Segmentation

引用

IEEE Access 2025年 13卷 83619-83642页

作者： AlZubi, Tareq Mahmod Mukhtar, Hamza Al-Balqa Applied University Prince Abdullah bin Ghazi Faculty of Information and Communication Technology Department of Computer Science As-Salt 19117 Jordan Department of Computer Science University of Engineering and Technology Lahore (UET Lahore) Lahore 39161 Pakistan Al-Khawarizmi Institute of Computer Science UET Lahore Intelligent Criminology Laboratory National Center of Artificial Intelligence Lahore 39161 Pakistan

The increasing reliance on medical image segmentation for disease diagnosis, treatment planning, and therapeutic assessment has highlighted the need for robust and generalized deep learning (DL)-based segmentation frameworks. However, existing models often suffer from task-specific limitations, catastrophic forgetting, and poor scalability due to their dependency on narrowly annotated datasets. This creates a significant gap in developing unified, multi-organ segmentation systems that leverage distributed and partially labeled datasets across diverse clinical institutions. To address these challenges, we propose the Federated 3D Knowledge Distillation Network (Fed3D-KDNet), a hybrid federated learning (FL) framework that integrates both global and local knowledge distillation mechanisms. Our model adapts the Segment Anything Model (SAM) for volumetric medical imaging by introducing architectural enhancements, including 3D spatial feature adapters and an Auto Prompt Generator (APG), to optimize spatial representation and reduce reliance on manually crafted prompts. Fed3D-KDNet employs a dual knowledge distillation strategy to mitigate catastrophic forgetting and improve cross-client knowledge transfer, ensuring robust generalization across heterogeneous datasets. The proposed methodology was evaluated on multi-organ CT datasets, including the BTCV benchmark, under centralized and federated settings. Experimental results demonstrate that Fed3D-KDNet achieves state-of-the-art performance with an average Dice score of 80.53% and an average Hausdorff Distance (HD) of 11.43 voxels in federated experiments involving seven clients, showing 5.04% improvement in Dice accuracy and a 4.35 voxel reduction in HD. Moreover, our model demonstrates superior efficiency with a computational cost of 371.3 GFLOPs, 26.53 million tuned parameters, and an inference time of 0.058 seconds per iteration. These results validate the efficacy, scalability, and computational efficiency of Fed3D-K

关键词： Federated learning knowledge distillation medical image segmentation transformer

来源：评论

学校读者我要写书评

暂无评论

Linearized Gait Phase Estimation Using BiLSTM

Linearized Gait Phase Estimation Using BiLSTM

引用

2025 International Conference on Electronics, Information, and Communication, ICEIC 2025

作者： Lee, Yonghyun Shin, Dongbin Ji, Younghoon Jang, Hyeyoun Han, Changsoo Lee, Yeonjoon Hanyang University Department Of Applied Artificial Intelligence Major In Bio Artificial Intelligence Seoul Korea Republic of RandD Center Hexar Humancare Ansan Korea Republic of Research and Development Hexar Humancare Ansan Korea Republic of CEO Hexar Humancare Ansan Korea Republic of Computer Science Hanyang University Ansan Korea Republic of

ISBN: (纸本)9798331510756

This paper presents a novel deep learning-based approach to address time-delay challenges arising from nonlinear transformations in Gait Phase Estimation, a critical factor in analyzing human gait for lower limb exoskeleton robots. The proposed model leverages hip joint angles and angular velocities as feature data, organized using a sliding window technique, to predict the Linearized Gait Phase (LGP)-a continuous repre-sentation of gait progression between Heel Strikes. Motivated by the need for real-time and precise gait phase estimation to enhance exoskeleton control, a Bidirectional LSTM network was employed to reduce delays and improve accuracy. The model's effectiveness was validated through comprehensive performance evaluations using metrics such as Mean Absolute Error (MAE), Root Mean Square Error (RMSE), Mean Square Error (MSE), and R2 score, alongside a temporal analysis of delay reduction. The results demonstrated superior performance, with an MAE of 2.8±0.8%, RMSE of 3.9±1.2%, MSE of 0.1±0.1%, and an R2 score of 98.1±1.8%. Particularly, temporal analysis revealed a marked improvement, achieving an MAE of 0%, compared to 2.74±0.92% reported in prior studies. These findings underscore the proposed model's potential for real-time gait phase estimation, offering significant implications for enhancing the responsiveness and adaptability of lower limb exoskeleton robots. © 2025 IEEE.

关键词： Exoskeleton (Robotics)

来源：评论

学校读者我要写书评

暂无评论

Some Results on Ricci-Yamabe Soliton and Its Cosmological Models

引用

IAENG International Journal of applied Mathematics 2025年第4期55卷 979-983页

作者： Pandey, Pankaj Yadav, Santosh Kumar Sharma, Kamakshi Chaturvedi, Braj Bhushan Department of Mathematics School of Chemical Engineering and Physical Sciences Lovely Professional University Punjab Phagwara144411 India Department of Mathematics School of Computer Science & Artificial Intelligence SR University Telangana Warangal506371 India Department of Pure & Applied Mathematics Guru Ghasidas Vishwavidyalaya Chhattisgarh Bilaspur495009 India

This article aims to investigate the characteristics of Ricci-Yamabe Soliton (briefly: (RYS)n). We study the cosmological models on (RYS)4 under Lorentzian para Sasakian (LPS)4 spacetime. Parallel Ricci tensor, Poisson structure and gradient property also reveal on (RYS)n. Example of (RYS)4 constructed to verify the results. © (2025), (International Association of Engineers). All rights reserved.

关键词： Solitons

来源：评论

学校读者我要写书评

暂无评论

Exploration of Diffusion-Based Test Case Generation of Autonomous Driving Systems

Exploration of Diffusion-Based Test Case Generation of Auton...

引用

International Conference on Electronics, Information and Communications (ICEIC)

作者： Joonwoo Lee Scott Uk-Jin Lee Department of Applied Artificial Intelligence Major in Bio-Artificial Intelligence Hanyang University Ansan Korea Department of Computer Science and Engineering Hanyang University Ansan Korea

ISBN: (数字)9798331510756

ISBN: (纸本)9798331510763

Autonomous driving systems demand extensive testing to guarantee safety and reliability. However, real-world testing is often costly and limited in variety. This paper investigates the use of diffusion models to generate synthetic driving images as a means of augmenting test datasets for autonomous driving systems. Leveraging their capability to produce diverse and highly realistic images, diffusion models can simulate various driving conditions, including rare or challenging scenarios that are difficult to replicate. By incorporating these synthetic images into the testing pipeline, self-driving systems can be assessed across a wider range of conditions, enhancing their robustness and safety.

关键词： Deep learning Technological innovation Text to image Transforms Diffusion models Data models Robustness Safety Autonomous vehicles Testing

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：