检索结果-内蒙古大学图书馆

Video and image quality improvement using an enhanced optimized dehazing technique

Multimedia Tools and Applications 2024年 1-19页

作者： Ayoub, Abeer El-Shafai, Walid El-Samie, Fathi E. Abd Hamad, Ehab K. I. EL-Rabaie, El-Sayed M. Electrical Engineering Department Faculty of Engineering Aswan University Aswan81542 Egypt Department of Electronics and Electrical Communication Engineering Faculty of Electronic Engineering Menoufia University Menouf32952 Egypt Security Engineering Lab Computer Science Department Prince Sultan University Riyadh11586 Saudi Arabia Department of Information Technology College of Computer and Information Sciences Princess Nourah Bint Abdulrahman University P.O. Box 84428 Riyadh11671 Saudi Arabia

The detrimental effects of atmospheric haze frequently plague outdoor imagery. This phenomenon arises from the scattering of light by minute particles within the ambient environment surrounding the scene to be imaged. Haze engenders an overall whitening of the image, leading to diminished contrast. To address these issues and enhance the quality of hazy images and videos, an enhanced dehazing technique is proposed. The proposed technique includes image enhancement before the optimal dehazing process. The enhancement stage entails the implementation of both homomorphic processing and Contrast Limited Adaptive Histogram Equalization (CLAHE), serving to control the image dynamic range, while concurrently heightening the image contrast. The culminating stage encompasses an optimized dehazing technique, adept at expunging haze-induced artifacts from images. The homomorphic processing and CLAHE, applied in the pre-processing step, establish a foundation for the subsequent dehazing procedure. This proposed methodology is systematically applied to a gamut of visual outputs, including visible videos, Near-Infrared (NIR) frames, and authentic hazy images. Comparative evaluations of the proposed technique, homomorphic-processing-based enhanced dehazing, and standalone dehazing techniques is undertaken on different video types encompassing five frames. It is evident that the proposed technique, synergizing the homomorphic processing, CLAHE, and dehazing, outperforms alternative strategies. Furthermore, the proposed technique is subjected to a comparative study with various existing dehazing techniques over real hazy images. The assessment depends on Peak Signal-to-Noise Ratio (PSNR), correlation, and entropy metrics. The results underscore the efficacy of the proposed technique, particularly in terms of spectral entropy enhancement of dehazed frames. For both visible and NIR frames, the percentages of enhancement by the proposed technique are 17.66% and 118.48%, respectively. ©

关键词： Image enhancement

来源：评论

学校读者我要写书评

暂无评论

Robust Multi-Dialect End-to-End ASR Model Jointly with Beam Search Threshold Pruning and LLM

引用

SN computer science 2025年第4期6卷 1-12页

作者： Shunmuga Priya, M.C. Karthika Renuka, D. Ashok Kumar, L. Department of Computer Science and Engineering Amrita School of Computing Amrita Vishwa Vidyapeetham Coimbatore India Department of Information Technology PSG College of Technology Coimbatore India Department of Electrical and Electronics Engineering Thiagarajar College of Engineering Madurai India

This paper aims to develop a novel robust multi-dialect end-to-end ASR system with beam search threshold pruning. The efficacy of our proposed model is evaluated using word error rate (WER). Our key contributions are: (1) To develop an end-to-end ASR system using attention-based neural network architecture and analyze the effectiveness of two features such as MFCC and log mel filter bank energies on multiple speech dialect corpora including American, Britain, and Indian accents;(2) To integrate beam search threshold pruning as a decoding mechanism to reduce the decoding time (3) To conduct an experimental analysis to test the model performance and compare the results against baseline system. (4) Post processing analysis are carried out using Llama2-7B based large language model(LLM) for enhancing the performance of proposed ASR system. The proposed model significantly improves performance by 1.91% and 4.29% over clean and noisy speech in librispeech corpus. Similarly, for the Indian accented speech, the model attains an average WER of about 6.6%. © The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd. 2025.

关键词： Automatic speech recognition Beam search Decoding LLM Log Mel filter bank energies

来源：评论

学校读者我要写书评

暂无评论

Path Planning for Self-driving Vehicles using Metaverse in 6G era with AI-enabled Networks 18

Path Planning for Self-driving Vehicles using Metaverse in 6...

引用

18th IEEE International Conference on Advanced Networks and Telecommunications Systems, ANTS 2024

作者： Choudhary, Rachit Prakash, Rahul Chaudhary, Rajat Budhiraja, Ishan Choudhary, Amit Sir Chhotu Ram Institute of Engineering and Technology Electronics and Communication Uttar Pradesh Meerut India Bennett University School of Computer Science and Engineering Technology Greater Noida India Kth Royal Institute of Technology Electrical Engineering and Computer Science Stockholm Sweden

ISBN: (纸本)9798350391725

Self-driving vehicles are the future of technology which is having the potential to reshape mobility by improving the safety, efficiency, and accessibility of transportation system in 6G networks. Critical task like safety is a crucial factor which is included in the planning of motions of the vehicle through an environment where other vehicles and other objects are part of analysis based on feedback control. This paper mainly focuses on decision-making for autonomous driving, specifically on lane change decisions by using deep reinforcement learning (DRL) in metaverse environment. The challenges of the lane change and path planning addresses the importance of considering both longitudinal speed and lateral lane changes in self-driving vehicle scenarios. The study defines the actions for the autonomous agent like either to stay in the current lane otherwise, changing lanes to the left and right. Thus, safety and efficiency are key concerns in the decision-making process based on rewards and penalties defined to encourage safe driving behavior. The paper limits the self-driving vehicle to drive in the three right lanes of the road to ensure safety and adherence to lane regulations. The goal is to drive as fast as possible within the speed limits, avoid collisions, and minimize unnecessary lane changes in self-driving robots environment. This paper implements longitudinal speed control through rule-based methods while focusing on the lane change decision-making process using reinforcement learning algorithms. Finally, the performance evaluation is performed on Matlab and Simulink to compute the path followed by the proposed model. © 2024 IEEE.

关键词： Motion planning

来源：评论

学校读者我要写书评

暂无评论

AI based Compact Controller for Wind farms Economical Operation, Control and Monitoring 2

AI based Compact Controller for Wind farms Economical Operat...

引用

2nd International Conference on Self Sustainable Artificial Intelligence Systems, ICSSAS 2024

作者： Sasikala, C. Palpandian, P. Muthuram, G. Kumar, S. Siva Karthik, S. Sam Raj, C. Mohan Arjun College of Technology Department of Electronics and Communication Engineering Coimbatore642 120 India Karpagam College of Engineering Department of Electrical and Electronics Engineering Coimbatore641 032 India Hindusthan College of Engineering and Technology Department of Electrical and Electronics Engineering Coimbatore India Nehru Institute of Engineering and Technology Department of Computer Science and Engineering Coimbatore641 105 India Dhanalakshmi Srinivasan College of Engineering Department of Electrical and Electronics Engineering Coimbatore641 105 India Sri Eshwar College of Engineering Department of Electrical and Electronics Engineering Coimbatore641 202 India

ISBN: (纸本)9798350368413

In India, 42.6 GW power is generated by wind turbine generators for supplying electricity to the load. Permanent Magnet DC generator, Synchronous generator and Induction generator are the most commonly used generators in the wind mill. Though the generators are designed for producing the power with maximum efficiency, monitoring, operation and control of wind mill plays an important role. The instantaneous faults like open circuit fault and least resistance fault can be easily detected and removed by the protection scheme. But the inter turn fault, core saturation condition, loss of dielectric due to ageing of machines cannot be detected until the fault occurs. The economic operation of the wind mill is based on the fault detection and prediction, uninterrupted operation of the wind mill, monitoring of parameters from the remote location. The AI based compact controller R04CPU with remote I/O module is used to monitor, control the windmill and ensures the economic operation/fault detection. © 2024 IEEE.

关键词： Windmill

来源：评论

学校读者我要写书评

暂无评论

Advanced Malware Detection Methods for Polymorphic Virus Identification 5

Advanced Malware Detection Methods for Polymorphic Virus Ide...

引用

5th IEEE International Conference on Communication, Computing and Industry 6.0, C2I6 2024

作者： Anil, D. Shreeshayana, R. Kiran, B. Preethi Uvce Bangalore University Department of Computer Science and Engineering Bengaluru India Atme College of Engineering Department of Electrical and Electronics Engineering Mysuru India Pes University Department of Computer Science and Engineering Bengaluru India Reva University School of Computing and Information Technology Bengaluru India

ISBN: (数字)9798350379884

ISBN: (纸本)9798350379884

Polymorphic viruses pose a significant challenge to traditional malware detection methods due to their ability to modify their code structure with each infection, effectively evading signature-based detection. This paper explores advanced malware detection methods tailored for identifying polymorphic viruses, focusing on the integration of machine learning, deep learning and hybrid approaches. Through a comprehensive analysis of static and dynamic features extracted from executables and URLs, this study employs classifiers such as Random Forests and Support Vector Machines (SVMs) to enhance detection accuracy. The SVM classifier demonstrated superior performance, achieving a detection accuracy of 96.23% compared to the 93.67% accuracy of the Random Forest classifier. This research highlights the potential of combining static and dynamic analysis with machine learning techniques to address the limitations of traditional methods, offering a robust framework for detecting and mitigating polymorphic and zero-day threats. © 2024 IEEE.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

VocalEyes: Enhancing Environmental Perception for the Visually Impaired through Vision-Language Models and Distance-Aware Object Detection

VocalEyes: Enhancing Environmental Perception for the Visual...

引用

2024 IEEE Conference on engineering Informatics, ICEI 2024

作者： Chavan, Kunal Balaji, Keertan Barigidad, Spoorti Chiluveru, Samba Raju Vellore Institute of Technology School of Electronics Engineering Vellore India S.R.M Institute of Technology Department of Computer Science and Engineering Chennai India Electrical Electronics and Communication Engineering IIT Dharwad Dharwad India

ISBN: (纸本)9798331505776

With an increasing demand for assistive technologies that promote the independence and mobility of visually impaired people, this study suggests an innovative real-time system that gives audio descriptions of a user's surroundings to improve situational awareness. The system acquires live video input and processes it with a quantized and fine-tuned Florence-2 big model, adjusted to 4-bit accuracy for efficient operation on low-power edge devices such as the NVIDIA Jetson Orin Nano. By transforming the video signal into frames with a 5 -frame latency, the model provides rapid and contextually pertinent descriptions of objects, pedestrians, and barriers, together with their estimated distances. The system employs Parler TTS Mini, a lightweight and adaptable Text-to-Speech (TTS) solution, for efficient audio feedback. It accommodates 34 distinct speaker types and enables customization of speech tone, pace, and style to suit user requirements. This study examines the quantization and fine-tuning techniques utilized to modify the Florence-2 model for this application, illustrating how the integration of a compact model architecture with a versatile TTS component improves real-time performance and user experience. The proposed system is assessed based on its accuracy, efficiency, and usefulness, providing a viable option to aid vision-impaired users in navigating their surroundings securely and successfully. © 2024 IEEE.

关键词： Speech enhancement

来源：评论

学校读者我要写书评

暂无评论

Smart Home Automation with Smart Metering using ZigBee Technology and Deep Belief Network 2

Smart Home Automation with Smart Metering using ZigBee Techn...

引用

2nd International Conference on Intelligent Data Communication Technologies and Internet of Things, IDCIoT 2024

作者： Murugesan, Sankar Cm, Avinash Bhuvanesh, A. Manoj, Sumitha Darney, P. Ebby Rajan S, Kani Siva Vel Tech Rangarajan Dr.Sagunthala RandD Institute of Science and Technology Department of Computer Science and Engineering Avadi Tamil Nadu Chennai India RajaRajeswari College of Engineering Department of Electrical and Electronics Engineering Karnataka Bangalore India PSN College of Engineering and Technology Department of Electrical and Electronics Engineering Tamil Nadu Tirunelveli India Rajarajeswari College of Engineering Department Electronics and Communication Engineering Karnataka Bengaluru India

ISBN: (纸本)9798350327533

The smart metering in the smart home application is vital one to ensure the reduction of energy usage and become a hot topic in the research field with the increasing usage of energy in both industrial and residential. With the advance technologies the energy saving techniques are used for the smart homes. The Internet of Things (IoT) devices are used for monitoring, controlling, and detecting the devices for the smart home applications. In this study, a novel Deep Belief Neural Network (DBN) approach for analyzing the data that are gathered from smart home applications with IoT devices, and big data technologies which promises the smart metering in smart home applications is proposed. The energy consumption patterns and classifications are effectuated with the proposed DBN approach and ensures the safety of the home. Experimental study was made and analyzed the performance of the proposed wo rk in terms energy consumption of different devices of smart home applications. Our proposed approach surpasses all the other approaches for the smart metering based smart applications. © 2024 IEEE.

关键词： Deep Belief Neural Network Energy Smart Home Smart Metering Zigbee Protocol

来源：评论

学校读者我要写书评

暂无评论

An Empirical Large-Signal Model of GaN HEMT for Design of High-Power Microwave Amplifier 12th

An Empirical Large-Signal Model of GaN HEMT for Design of Hi...

引用

12th International Conference on Frontiers of Intelligent Computing: Theory and Applications, FICTA 2024

作者： Sharma, Swati Alsabbagh, Moh. Yagoub, Mustapha C. E. Sharma, Shikha Swaroop Pandey, Atul Kumar Department of Electronics Engineering University of Rome Tor Vergata Rome00133 Italy School of Electrical Engineering and Computer Science University of Ottawa OttawaON Canada New Delhi110031 India

ISBN: (纸本)9789819601424

This work described a well-organized empirical modelling approach of large-signal GaN HEMT equivalent circuit for high-power applications. All the stages involved in the extraction have been described in this paper, beginning with the determination of extrinsic elements, and finishing with the intrinsic ones. Nonlinear capacitances and resistances have been extracted and modelled for saturation region, sub-threshold region as well as the triode region. Only six parameters were used to make the new channel current model, which represented the unique working principles of HEMT. The self-heating effect, virtual gate formation and dispersion phenomena have been integrated in the channel current model. As, only DC and S-parameters data have been required, this model has reasonably simple to develop and use in CAD software. The nonlinear features of the device and the harmonic components have been properly predicted by the proposed model. Small-signal to large-signal excitation regions showed excellent agreements using ADS software. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Gallium nitride

来源：评论

学校读者我要写书评

暂无评论

A Novel Technique to Improve Latency and Response Time of AI Models using Serverless Infrastructure 6

A Novel Technique to Improve Latency and Response Time of AI...

引用

6th International Conference on Inventive Computation Technologies, ICICT 2023

作者： Ravi, Bhanu Sankar Madhukanthi, Chepuri Sivasankar, P. Prasanna, John Deva Department of Computer Science Engineering Hindustan Institute of Technology and Science Chennai India Department of Electrical Electronics and Communication NITTTR Chennai India

ISBN: (纸本)9798350398496

Response time for AI-based models in time-critical applications is always a matter of concern. The situation is challenging when the model is deployed in a cloud-based infrastructure. To address this issue, a cloud-native development methodology called serverless infrastructure enables developers to create and execute applications without having to worry about managing servers. Generally, traditional systems require manual scaling, constant maintenance, and dedicated hardware resulting in high costs. To solve this, the AI model is deployed in serverless infrastructure services for hosting APIs that is to identify marine animals which have the potential to attack human beings at the seashores. The serverless infrastructure suffers from an initializing delay called cold start for occasional requests and hence the response time will be delayed even if the lambda function is free. The problem of cold starts is mitigated using the scheduler in the lambda function. The scheduler sends dummy requests to the server to keep the server warm and active. The AI model used as a test case utilizes Convolutional Neural Network Algorithms and Transfer Learning Technique, for detecting predators near the seashore. The model is deployed in a serverless infrastructure and has the benefits of automatic scaling, pay-per-use pricing, and decreased operational costs. The AI model is implemented in both serverless infrastructure and in Elastic Cloud Compute EC2. The performance of both systems was done for cost, latency, and response time. The proposed system provides promising results when compared to traditional server systems. © 2023 IEEE.

关键词： Image classification

来源：评论

学校读者我要写书评

暂无评论

Medical knowledge graph question answering for drug-drug interaction prediction based on multi-hop machine reading comprehension

引用

CAAI Transactions on Intelligence Technology 2024年第5期9卷 1217-1228页

作者： Peng Gao Feng Gao Jian-Cheng Ni Yu Wang Fei Wang Qiquan Zhang School of Cyber Science and Engineering Qufu Normal UniversityQufuChina School of Computer Science and Technology East China Normal UniversityShanghaiChina Network and Information Center Qufu Normal UniversityQufuChina School of Electronics and Information Engineering Harbin Institute of TechnologyShenzhenChina School of Electrical Engineering and Telecommunications University of New South WalesSydneyNSWAustralia

Drug-drug interaction(DDI)prediction is a crucial issue in molecular *** methods of observing drug-drug interactions through medical experiments require significant resources and *** authors present a Medical Knowledge Graph Question Answering(MedKGQA)model,dubbed MedKGQA,that predicts DDI by employing machine reading comprehension(MRC)from closed-domain literature and constructing a knowledge graph of“drug-protein”triplets from open-domain *** model vectorises the drug-protein target attributes in the graph using entity embeddings and establishes directed connections between drug and protein entities based on the metabolic interaction pathways of protein targets in the human *** aligns multiple external knowledge and applies it to learn the graph neural *** bells and whistles,the proposed model achieved a 4.5%improvement in terms of DDI prediction accuracy compared to previous state-of-the-art models on the QAngaroo MedHop *** results demonstrate the efficiency and effectiveness of the model and verify the feasibility of integrating external knowledge in MRC tasks.

关键词： artificial neural network data analysis document handling natural language processing

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：