检索结果-内蒙古大学图书馆

engineering & Computing Technologies (ICECT), International Conference on

作者： Nuaima Saeed Department of Artificial Intelligence Faculty of Computer Science & Engineering Ghulam Ishaq Khan Institute of Engineering Sciences and Technology Topi Pakistan

ISBN: (数字)9798350349719

ISBN: (纸本)9798350349726

Generative AI and text-based image synthesis are being pursued with more interest than ever and adopt a growing audience. This work offers a new way of producing images within a broad range of quality levels, deducing the extra value of the increase resolution. The system applies sophisticated technologies such as diffusion and upscaling models that are guided by the pre-training of the Contrastive Language-Image PreTraining (CLIP). However, the model shines primarily in its ability to generate visually cohesive and artistic images which are a good illustration of each specified textual prompt. Upsampling technique resolved the major translation problems by generating more high resolution and detailed resultant images of base images. Beyond that, the system also thoroughly illustrated its diversity by accurately producing a comprehensive collection of images based on the text input. This work would largely be a demonstration of the suitability of it for different fields like digital art, advertising, and content creation. Moreover, it delimits the chances of producing innovative and unique content. Future development of many other applications of the system will include the improvement of the system handling more complex textual descriptions, the addition of user-interactive generation features, the optimization of system performance and expanding to multimodal content generation. Our results here only represent a small but important step in this exciting direction. By means of this survey we enrich further knowledges on the field and its farther issues.

关键词： Surveys Image resolution Image synthesis Generative AI system performance Text to image Digital art

来源：评论

学校读者我要写书评

暂无评论

Latent Communication in Artificial Neural Networks

arXiv

引用

arXiv 2024年

作者： Moschella, Luca GLADIA research lab Department of Computer Science Faculty of Information Engineering Informatics and Statistics Italy

As NNs (Neural Networks) permeate various scientific and industrial domains, understanding the universality and reusability of their representations becomes crucial. At their core, these networks create intermediate neural representations, indicated as latent spaces, of the input data and subsequently leverage them to perform specific downstream tasks. This dissertation focuses on the universality and reusability of neural representations. Do the latent representations crafted by a NN remain exclusive to a particular trained instance, or can they generalize across models, adapting to factors such as randomness during training, model architecture, or even data domain? This adaptive quality introduces the notion of Latent Communication – a phenomenon that describes when representations can be unified or reused across neural spaces. A salient observation from our research is the emergence of similarities in latent representations, even when these originate from distinct or seemingly unrelated NNs. By exploiting a partial correspondence between the two data distributions that establishes a semantic link, we found that these representations can either be projected into a universal representation (Moschella*, Maiorca*, et al., 2023), coined as Relative Representation, or be directly translated from one space to another (Maiorca* et al., 2023). Intriguingly, this holds even when the transformation relating the spaces is unknown (Cannistraci, Moschella, Fumero, et al., 2024) and when the semantic bridge between them is minimal (Cannistraci, Moschella, Maiorca, et al., 2023). Latent Communication allows for a bridge between independently trained NN, irrespective of their training regimen, architecture, or the data modality they were trained on – as long as the data semantic content stays the same (e.g., images and their captions). This holds true for both generation, classification and retrieval downstream tasks;in supervised, weakly supervised, and unsupervised settings;and

关键词： Neural networks

来源：评论

学校读者我要写书评

暂无评论

Monkeypox Lesion Classification: A Transfer Learning Approach for Early Diagnosis and Intervention 7

Monkeypox Lesion Classification: A Transfer Learning Approac...

引用

7th International Conference on Contemporary Computing and Informatics, IC3I 2024

作者： Al Noman, Abdullah Fardin, Hasib Chhabra, Gunjan Sultana, Shayma Haque, Rezaul Ahmed, Md. Redwan Hasibunnahar, Swarna Kaushik, Keshav Westcliff University Department of Information Technology United States Westcliff University Department of Engineering Management United States Graphic Era Hill University Department of Cse Uttarakhand Dehradun India Graphic Era Deemed to be University Uttarakhand Dehradun India International American University Department of Management Information System United States East West University Department of Computer Science and Engineering Bangladesh Huzhou University Department of Life Sciences China Amity University Amity School of Engineering and Technology Punjab Mohali India

ISBN: (纸本)9798350350067

Monkeypox is a viral disease found in Central and West Africa, and it presents notable public health challenges due to its clinical similarity to smallpox. Recent outbreaks in over 40 countries have underscored the urgent need for quick and accurate diagnosis, especially in settings with limited resources where traditional diagnostic methods may be inadequate. This study explores the application of deep learning techniques to classify four specific types of monkeypox skin lesions. Transfer learning was employed to evaluate the performance of different models using a dataset of 770 skin lesion images. Various methods, including data augmentation and image preprocessing, were utilized to improve the effectiveness of the models. The results indicate that the custom SqueezeNet model was the most successful, achieving an accuracy of 98.85% and an F1 score of 98.89%. These results surpass previous research benchmarks across four key evaluation measures. Additionally, the model was integrated into a web-based diagnostic tool capable of reliably detecting different forms of monkeypox. This study represents a significant advancement in utilizing artificial intelligence for disease detection, offering a practical approach for early and accurate monkeypox diagnosis. The proposed system has thex potential for seamless integration into clinical practice, aiding healthcare professionals in timely interventions. © 2024 IEEE.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Influencing Factors in Facial Recognition and Verification: A Siamese Neural Network Hyperparameter Study

Influencing Factors in Facial Recognition and Verification: ...

引用

engineering & Computing Technologies (ICECT), International Conference on

作者： Ahmad Maaz Usama Arshad Tafreed Ahmad Shaheer Abbas Department of Artificial Intelligence Faculty of Computer Science & Engineering Ghulam Ishaq Khan Institute of Engineering Sciences and Technology Topi Pakistan

ISBN: (数字)9798350349719

ISBN: (纸本)9798350349726

There are many practical uses for facial verification and recognition jobs. In this study, we look into the developments and difficulties faced in facial recognition technologies, especially in the application of Siamese neural networks. This study converges around optimizing hyperparameters to improve Siamese network accuracy and performance. We evaluated the impact of different optimizers, loss functions, and data preprocessing techniques on Siamese network performance through a chain of experiments. The results demonstrated the importance of choosing the right hyperparameters, demonstrating how decisions such as the Adam optimizer and Binary Cross-Entropy as the loss function may greatly enhance the system's capability to precisely identify and validate faces. This study also overlooks the importance of diverse and high-quality datasets and states how these elements affect the model's performance. With this work, we hope to give a thorough manual for fine-tuning Siamese neural networks for use in facial recognition applications, along with helpful advice on how to overcome the difficulties posed by dataset variability and hyperparameter selection. In addition to adding to the theoretical knowledge of facial verification technologies, our research offers useful suggestions for enhancing the conception and application of these systems in actual settings.

关键词： Training computers Accuracy Face recognition Neural networks Data preprocessing Manuals

来源：评论

学校读者我要写书评

暂无评论

Optimized Control Strategy for LC L-Filtered Three-Phase Voltage-Source Converters in Adverse Grid Conditions

Optimized Control Strategy for LC L-Filtered Three-Phase Vol...

引用

IEEE Industry Applications Society Annual Meeting (IAS)

作者： Ahmad Ali Nazeri Jens Friebe Peter Zacharias Department of Power Electronics Faculty of Electrical Engineering / Computer Science University of Kassel Kassel Germany

ISBN: (数字)9798350372717

ISBN: (纸本)9798350372724

This paper presents the precise AC current control within three-phase grid-connected voltage source converters featuring an $L C L$ filter, particularly in distributed generation systems. Traditional control methodologies such as proportionalintegral (PI) and proportional resonant (PR) are conventionally employed to ensure zero steady-state error at the fundamental frequency. However, to effectively address high-order harmonic currents, this study introduces the utilization of proportional resonant plus multi-resonant (PR-MR) techniques within the current control loop to mitigate current harmonics induced by dead time and distorted grid voltages. The paper offers a comprehensive current control design methodology, incorporating a rigorous system stability analysis. Experimental findings demonstrate a significant reduction in total harmonic distortion of grid currents, achieving approximately 1.1 % under distorted grid conditions, thereby meeting the stringent IEEE 1547 standard. The proposed controllers are implemented on the TMDSDOCK28379D 32bit floating-point digital signal processor (DSP). The system's performance is evaluated through simulations conducted in the MATLAB/Simulink environment and validated experimentally on a 7.5 kW three-phase $L C L$ -filtered grid-connected VSC employing the proposed control strategies.

关键词： Current control Total harmonic distortion system performance Digital signal processors Harmonic analysis Steady-state Voltage control Power harmonic filters Standards Voltage-source converters

来源：评论

学校读者我要写书评

暂无评论

Experiment Design and Verification for Assessing the Acquisition of Strategic Planning Ability

Experiment Design and Verification for Assessing the Acquisi...

引用

Advanced Applied Informatics Winter (IIAI-AAI-Winter), International Congress on

作者： Natchanon Manatphaiboon Shogo Hamachi Zeynep Yücel Pattara Leelaprute Akito Monden Department of Computer Engineering Faculty of Engineering Kasetsart University Bangkok Thailand Department of Computer Science Okayama University Okayama Japan

Exposing healthy subjects to the Tower of London (ToL) task several times enables the observation of the evolution of skill level over time as well as its reflections on eye gaze patterns. In order to identify such effects, one needs a carefully designed set of ToL tasks (i.e. problem sets). In particular, problem complexity should be gradually increasing within each problem set and comparable across different problem sets, as assessed in terms of objective measures, to enable smooth progress and the accumulation of experience. In that respect, this study uses goal hierarchy, search depth, and the number of optimal solutions for assessing complexity and establishing the gradual relation. Namely, two problem sets, Base-1 and Base-2, are designed. To satisfy the comparability condition, six color permutations are derived from Base-1 (experiments-1 to -6) and delivered to the subjects on two consecutive days. Finally, another problem set derived from Base-2 with the exact same structural properties is delivered on the last day (experiment-7). We conducted experiments with 15 participants and examined the evolution of their planning, trial, and execution times, as well as rule violations and the number of moves beyond optimal. The results suggest that there are indeed improvements in speed and accuracy in experiment-7 compared to experiment-1, even though both are novel to the participants. Additionally, these improvements are shown to be similar and consistent for different (initial) problem sets. Moreover, the proposed problem sets' difficulty and planning time for all experiments have a strong positive correlation, averaging $0.86 (\sigma=0.06)$ . In that respect, we believe that the eye-gaze patterns also have the potential to withhold important information and are worth further investigation in future studies.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Practical Acoustic Eavesdropping On Typed Passphrases

arXiv

引用

arXiv 2025年

作者： Fürst, Darren Aßmuth, Andreas Department of Electrical Engineering Media and Computer Science Ostbayerische Technische Hochschule Amberg-Weiden Amberg Germany Faculty of Computer Science and Electrical Engineering Kiel University of Applied Sciences Kiel Germany

Cloud services have become an essential infrastructure for enterprises and individuals. Access to these cloud services is typically governed by Identity and Access Management systems, where user authentication often relies on passwords. While best practices dictate the implementation of multi-factor authentication, it’s a reality that many such users remain solely protected by passwords. This reliance on passwords creates a significant vulnerability, as these credentials can be compromised through various means, including side-channel attacks. This paper exploits keyboard acoustic emanations to infer typed natural language passphrases via unsupervised learning, necessitating no previous training data. Whilst this work focuses on short passphrases, it is also applicable to longer messages, such as confidential emails, where the margin for error is much greater, than with passphrases, making the attack even more effective in such a setting. Unlike traditional attacks that require physical access to the target device, acoustic side-channel attacks can be executed within the vicinity, without the user’s knowledge, offering a worthwhile avenue for malicious actors. Our findings replicate and extend previous work, confirming that cross-correlation audio preprocessing outperforms methods like mel-frequency-cepstral coefficients and fast-fourier transforms in keystroke clustering. Moreover, we show that partial passphrase recovery through clustering and a dictionary attack can enable faster than brute-force attacks, further emphasizing the risks posed by this attack vector. Copyright © 2025, The Authors. All rights reserved.

关键词： Unsupervised learning

来源：评论

学校读者我要写书评

暂无评论

Development and Validation of Customer Intention Factors for Live-Stream Shopping 7

Development and Validation of Customer Intention Factors for...

引用

7th International Conference on New Media Studies, CONMEDIA 2023

作者： Fernando, Erick Hidayat, Desman Pengaribuan, Christian Haposan Prabowo, Yulius Denny Universitas Multimedia Nusantara Faculty of Engineering and Informatics Department of Information System Tangerang Indonesia Bina Nusantara University Binus Entrepreneurship Center Management Department Jakarta11480 Indonesia Sampoerna University Management Study Program Faculty of Business Jakarta12780 Indonesia Bina Nusantara University Binus Online Learning Computer Science Department Jakarta11480 Indonesia

ISBN: (纸本)9798350307504

Along with technological improvements, online shopping is currently developing quickly. Online shops started by selling electronics, clothing, food, and home appliances and continue to evolve, selling various things. Currently, digital marketing has a novel technique, namely live shopping, where sellers can present, promote and offer their products directly through live streaming on social media and e-commerce platforms. This research will focus on gaining insight into the factors influencing customers to watch and purchase through live streaming. This research is an early-stage study of exploring and searching for research models. The study used a literature review approach by reviewing similar previous research articles. Several variable components were used in this study. The findings of the factors include visibility capacity, meta-vocation capacity, purchase orientation capacity, social presence, price, perceived product quality, perceived enjoyment, purchase orientation, perceived usefulness, and intention to buy. These factors were tested and proved to be valid with Cronbach's Alpha value being more than 0.5 and reliable with the value of the loading factor a greater than 0.6. Criteria The factors of the findings for further analysis of the relationship between factors and other variables. © 2023 IEEE.

关键词： Domestic appliances

来源：评论

学校读者我要写书评

暂无评论

Bi-Directional Power Control in Grid-Connected Electric Vehicle On-Board Chargers using Spider Wasp Optimization

Bi-Directional Power Control in Grid-Connected Electric Vehi...

引用

Trends in Material science and Inventive Materials (ICTMIM), International Conference on

作者： Cuddapah Anitha G Vipashi Kansal Swapnil Uttamrao Deokar Kamal Sutaria Ravi Kumar V Natrayan L Department of Computer Science and Engineering School of Computing Mohan Babu University Tirupati Andhra Pradesh Department of Computer Science & Engineerin Graphic Era Deemed to be University Dehradun Department of Mechanical Engineering Smt. Kashibai Navale College of Engineering Pune Computer Science and Engineering Faculty of Engineering and Technology Parul Institute of Engineering and Technology Parul University Gujarat Department of Mechanical Engineering Amrita School of Engineering Amrita Vishwa Vidyapeetham Bengaluru Department of Mechanical Engineering Saveetha School of Engineering SIMATS Chennai Tamil Nadu

ISBN: (数字)9798331501488

ISBN: (纸本)9798331501495

The bi-directional power control in electric vehicle (EV) on-board chargers aims to maximize charging performance while minimizing size and cost of the system. This presents an increasing technical challenge that further complication of thermal management through greater integration of power conversion components may impact system reliability, efficiency, and longevity. By address these issues this paper proposes a Spider Wasp Optimization (SWO) for enhancing stability in bi-directional power control in EV on-board chargers. The objective is to minimize Total Harmonic Distortion (THD) in EV on-board chargers. By optimizing the converter's parameters, SWO effectively reduces harmonic distortions, leading to improved power quality and smoother operation of bi-directional power control in EV on-board chargers. By then the proposed SWO method is implemented in MATLAB platform and evaluated their performance with various existing methods such as Improved Particle Swarm Optimization (IPSO), Intelligent Neural Herd based Water Drop Optimization (INHWDO), Squirrel Wolf Optimization algorithm (SWOA), Wild Horse-Particle Swarm Optimization (WHPSO) and Assailant Inspired Chimp Optimization Algorithm (AICOA). The proposed SWO method outperforms all the others with the lowest THD of 2.008%, indicating its superior performance in reducing harmonic distortion.

关键词： Total harmonic distortion Power control Power quality Bidirectional control Electric vehicles Real-time systems Vehicle dynamics Particle swarm optimization Optimization Thermal stability

来源：评论

学校读者我要写书评

暂无评论

Adaptive Memory Differential Evolutionary Algorithm for Network Management Using Variants of Dominating Set Models

Adaptive Memory Differential Evolutionary Algorithm for Netw...

引用

Intelligent Methods, systems, and Applications (IMSA)

作者： Abdel-Rahman Hedar Fatma El-Zahraa Tarek Mostafa Mohamed Gamal A. El-Sayed Department of Computer Science Faculty of Computers & Information Assiut University Assiut Egypt Department of Electrical Engineering Faculty of Engineering Assiut University Assiut Egypt Department of Computer Science in Jamoum Umm Al-Qura University Makkah Saudi Arabia

ISBN: (数字)9798350362633

ISBN: (纸本)9798350362640

In wireless sensor networks, where multiple sensors are typically concentrated in a confined area, determining the optimal size of wireless sensors to use for communication and coordination over the network is essential due to concerns about the preservation of energy and battery. In this paper, we introduce new competitive memory-based differential evolution methods to solve wireless sensor network management problems using variants of dominating set models. The problems considered are formulated as constrained optimization problems that are transformed into unconstrained optimization problems using the penalty principle. Then, differential evolution with several memory-based enhanced procedures is designed to find solutions to the reformed optimization problems. Two versions of the proposed Adaptive Memory Differential Evolutionary Algorithm (AMDEA) are designed to deal with the minimum dominating set problem and its variants. The results obtained from the proposed methods show that they are promising compared to some benchmark methods.

关键词： Wireless communication Wireless sensor networks Adaptation models Memory management Evolutionary computation Benchmark testing Sensor systems and applications

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：