检索结果-内蒙古大学图书馆

A transfer learning approach for continuous speech recognition system in Indian language sadri

International Journal of Information technology (Singapore) 2025年 1-10页

作者： Paul, Shubhojeet Bhattacharjee, Vandana Saha, Sujan Kumar Department of Computer Science and Engineering Birla Institute of Technology Mesra Jharkhand Ranchi 835215 India Department of Computer Science and Engineering National Institute of Technology West Bengal Durgapur 713209 India

Sadri is the most widely used language of the Chotanagpur Plateau region of India. This is primarily a spoken language and developing an automatic speech recognition (ASR) system in Sadri is extremely important. When we searched the literature, we found no open ASR system or relevant resources in Sadri. So, we worked on developing an ASR system in Sadri. We created a Sadri speech corpus of around 20 h, employing 53 native speakers. Then, we implemented the baseline model using Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN). As the size of the training data was not sufficient, we investigated a transfer learning technique to improve performance. We found that ASR data are openly available in other related languages such as Hindi and Bengali. We used Bengali and Hindi data of 100 h each in a transfer learning framework along with the Sadri data. In our experiments, we found that the transfer learning-based model outperforms the baseline model. © Bharati Vidyapeeth's Institute of computer Applications and Management 2025.

关键词： ASR Deep learning Indian language Sadri Speech recognition

来源：评论

学校读者我要写书评

暂无评论

A Novel Fuzzy-Based Modified GAN and Faster RCNN for Classification of Banana Leaf Disease

引用

Journal of The Institution of Engineers (India): Series A 2023年第3期104卷 529-540页

作者： Raja, N. Bharathi Rajendran, P. Selvi Department of Computer Science and Engineering Hindustan Institute of Technology and Science Tamil Nadu Kelambakkam India

In the global marketplace, agriculture plays an important role. However, diseases produced in plants mostly affect the financial system. Pest occurrence and climatic changes are the leading trouble that the banana plant leaf faces, which influences the quality of fruit and crop yields. To predict the leaf disease symptoms in an early stage, the automatic banana leaf disease detection technique is more significant. The traditional deep neural network methods used in banana plant leaf detection need more parameters and utilize complex architectures. To address such problems, modified generative adversarial networks–modified faster region-based convolutional neural networks with fuzzy (MGAN–MFRCNN with Fuzzy) is proposed for identifying banana leaf disease. The most effective framework in deep learning is GAN. The autoencoder is incorporated within the fundamental GAN design for developing a modified GAN model;it is used for data augmentation. To perform feature extraction from high-resolution networks, an MFRCNN is used in our proposed method. The improved and more accurate classification scores are obtained by combining fuzzy with the ensemble model. The experimentation is carried out using a banana plant leaf dataset in which the healthy leaf, xanthomonas-affected leaf, and sigatoka-affected leaf diseases are detected and classified. The proposed model reached 98% accuracy, 97% precision, 96% F1-score, 98% specificity, and 98% sensitivity, which is better than other existing methods. © 2023, The Institution of Engineers (India).

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Mixed-decomposed convolutional network:A lightweight yet efficient convolutional neural network for ocular disease recognition

引用

CAAI Transactions on Intelligence technology 2024年第2期9卷 319-332页

作者： Xiaoqing Zhang Xiao Wu Zunjie Xiao Lingxi Hu Zhongxi Qiu Qingyang Sun Risa Higashita Jiang Liu Research Institute of Trustworthy Autonomous Systems and Department of Computer Science and Engineering Southern University of Science and TechnologyShenzhenChina Tomey Corporation NagoyaJapan Guangdong Provincial Key Laboratory of Brain‐inspired Intelligent Computation Department of Computer Science and EngineeringSouthern University of Science and TechnologyShenzhenChina Singapore Eye Research Institute SingaporeSingapore

Eye health has become a global health concern and attracted broad *** the years,researchers have proposed many state-of-the-art convolutional neural networks(CNNs)to assist ophthalmologists in diagnosing ocular diseases efficiently and ***,most existing methods were dedicated to constructing sophisticated CNNs,inevitably ignoring the trade-off between performance and model *** alleviate this paradox,this paper proposes a lightweight yet efficient network architecture,mixeddecomposed convolutional network(MDNet),to recognise ocular *** MDNet,we introduce a novel mixed-decomposed depthwise convolution method,which takes advantage of depthwise convolution and depthwise dilated convolution operations to capture low-resolution and high-resolution patterns by using fewer computations and fewer *** conduct extensive experiments on the clinical anterior segment optical coherence tomography(AS-OCT),LAG,University of California San Diego,and CIFAR-100 *** results show our MDNet achieves a better trade-off between the performance and model complexity than efficient CNNs including MobileNets and ***,our MDNet outperforms MobileNets by 2.5%of accuracy by using 22%fewer parameters and 30%fewer computations on the AS-OCT dataset.

关键词： artificial intelligence deep learning deep neural networks image analysis image classification medical applications medical image processing

来源：评论

学校读者我要写书评

暂无评论

Evolutionary recurrent neural network based on equilibrium optimization method for cloud-edge resource management in internet of things

引用

Neural Computing and Applications 2025年第6期37卷 4957-4969页

作者： Ebrahimi Mood, Sepehr Rouhbakhsh, Adel Souri, Alireza Department of Computer Science Yazd University Yazd Iran Department of Computer Engineering Faculty of Engineering Haliç University Istanbul34060 Turkey School of Electrical and Electronic Engineering Shandong University of Technology Zibo255000 China

Cloud computing is an emerging field in information technology, enabling users to access a shared pool of computing resources. Despite its potential, cloud technology presents various challenges, with one of the most significant being resource management within the cloud environment. Specifically, the allocation of physical machines (PMs) to virtual machines (VMs) in the Infrastructure as a service (IaaS) layer aims to achieve optimal efficiency. To address this challenge, we propose an evolutionary recurrent neural network (RNN) for resource allocation, enhanced by the equilibrium optimization (EO) algorithm. This method improves the generalization performance and reduces the architecture complexity of RNNs, while ensuring better adaptability to various demand patterns in cloud environments. Additionally, the effectiveness of RNNs heavily relies on the quality of training data;hence, we generate appropriate training data to facilitate the development of an efficient resource management model. Through extensive simulations, our method demonstrates superior performance compared to traditional algorithms. However, it is worth noting that this study is primarily based on simulation results, highlighting the need for real-world validation to confirm its practical applicability. The proposed method's applicability to different cloud computing environments and scenarios, including various service models (IaaS, PaaS, and SaaS) and deployment models (public, private, and hybrid), is also briefly discussed. Future studies should explore the computational efficiency and potential optimization techniques of the EO-RNN approach and consider real-world implementations to validate its effectiveness and uncover practical challenges. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024.

关键词： Cloud platforms

来源：评论

学校读者我要写书评

暂无评论

Image Forgery Detection Model using CNN Architecture with SVM Classifier 7

Image Forgery Detection Model using CNN Architecture with SV...

引用

7th International Conference on Parallel, Distributed and Grid Computing, PDGC 2022

作者： Singh, Surjeet Sehgal, Vivek Kumar Jaypee University of Information Technology Department of Computer Science and Engineering and Infotmation Technology Waknaghat Himachal Pradesh Solan India

ISBN: (纸本)9781665454018

This Due to rapid advancement in computer vision system and easy availability of digital camera's acquisition of images and sharing the content on social media platform with falsify content is challenging task to the researcher community. For the legal purpose to provide the evidence in court, it's a difficult task identify the forgery in image that why various traditional artefacts of capturing device and internal footprints of image used to check the forgery in original image. Now a day's easy availability of software editing tools the captured image modify with wrong content to defame the person or misleading the people on social media with fake content. In this article, proposed a CNN based architecture to classify the forgery in given image. CNN architecture is capable to detect the unseen forgeries based on features extracted through various convolution layers and used SVM classifier to classify the label of forged image with high accuracy. © 2022 IEEE.

关键词： Digital forensics

来源：评论

学校读者我要写书评

暂无评论

Formalization of the Telegrapher’s Equations using Higher-Order-Logic Theorem Proving

Journal of Applied Logics

引用

Journal of Applied Logics 2024年第2期11卷 197-236页

作者： Deniz, Elif Rashid, Adnan Hasan, Osman Tahar, Sofiène Department of Electrical and Computer Engineering Concordia University MontrealQC Canada School of Electrical Engineering and Computer Science National University of Sciences and Technology Islamabad Pakistan

The telegrapher’s equations constitute a set of linear partial differential equations that establish a mathematical correspondence between the electrical current and voltage within transmission lines, taking into account factors, such as distance and time. These equations find wide applications in the design and analysis of various systems, including integrated circuits and antennas. This paper proposes the utilization of higher-order-logic theorem proving for a formal analysis of the telegrapher’s equations, also referred to as the transmission line equations. Specifically, we present a formal model of the telegrapher’s equations in both time and phasor domains. Subsequently, we employ the HOL Light theorem prover to formally verify the solutions of the telegrapher’s equations in the phasor domain. Furthermore, we established a connection between phasor and time-domain functions to formally verify the general solutions for the time-domain partial differential equations for the current and voltage in an electric transmission line. To demonstrate the practical effectiveness of our proposed formalization, we conduct a formal analysis of a terminated transmission line and its special cases, i.e., short-and open-circuited transmission lines commonly used in antenna design, by formally verifying the load impedance and the voltage reflection coefficient. © 2024, College Publications. All rights reserved.

关键词： Electric lines

来源：评论

学校读者我要写书评

暂无评论

EAMR-Net: A multiscale effective spatial and cross-channel attention network for retinal vessel segmentation

引用

Mathematical Biosciences and engineering 2024年第3期21卷 4742-4761页

作者： Prethija, G. Katiravan, Jeevaa School of Computer Science and Engineering Vellore Institute of Technology Chennai600127 India Department of Information Technology Velammal Engineering College Chennai600066 India

Delineation of retinal vessels in fundus images is essential for detecting a range of eye disorders. An automated technique for vessel segmentation can assist clinicians and enhance the efficiency of the diagnostic process. Traditional methods fail to extract multiscale information, discard unnecessary information, and delineate thin vessels. In this paper, a novel residual U-Net architecture that incorporates multi-scale feature learning and effective attention is proposed to delineate the retinal vessels precisely. Since drop block regularization performs better than drop out in preventing overfitting, drop block was used in this study. A multi-scale feature learning module was added instead of a skip connection to learn multi-scale features. A novel effective attention block was proposed and integrated with the decoder block to obtain precise spatial and channel information. Experimental findings indicated that the proposed model exhibited outstanding performance in retinal vessel delineation. The sensitivities achieved for DRIVE, STARE, and CHASE_DB datasets were 0.8293, 0.8151 and 0.8084, respectively. © 2024the Author(s).

关键词： Drops

来源：评论

学校读者我要写书评

暂无评论

A Comprehensive Study: Image Forensic Analysis Traditional to Cognitive Image Processing 8

A Comprehensive Study: Image Forensic Analysis Traditional t...

引用

8th International Conference on Signal Processing and Communication, ICSC 2022

ISBN: (纸本)9781665454308

Image and video forensics is the biggest challenge in current digital era due to rapid change or modification in digital content by using lots of available free software and editing tools. Authenticity of image and video is explained from image acquisition phase to storage phase through various forensics methods. The aim of this survey is to investigate image and video forensics methods using the intrinsic footprints of digital image during the entire image processing life cycle in the traditional image processing paradigm to cognitive image processing paradigm. © 2022 IEEE.

关键词： Life cycle

来源：评论

学校读者我要写书评

暂无评论

Performance Analysis of ZF and RZF in Low-Resolution ADC/DAC Massive MIMO Systems

引用

China Communications 2024年第8期21卷 115-126页

作者： Talha Younas Shen Jin Muluneh Mekonnen Gao Mingliang Saqib Saleem Sohaib Tahir Mahrukh Liaqat School of Electrical and Electronic Engineering Shandong University of TechnologyZibo 255000China Department of Electrical and Computer Engineering COMSATS University IslamabadSahiwal Campus Department of Electrical and Computer Engineering College of EngineeringDhofar UniversitySalalah 211Oman Department of Electrical and Computer Engineering Addis Ababa Science and Technology UniversityAddis AbabaEthiopia Department of Electrical Engineering College of Electrical and Mechanical EngineeringNUSTIslamabadPakistan

Large number of antennas and higher bandwidth usage in massive multiple-input-multipleoutput(MIMO)systems create immense burden on receiver in terms of higher power *** power consumption at the receiver radio frequency(RF)circuits can be significantly reduced by the application of analog-to-digital converter(ADC)of low *** this paper we investigate bandwidth efficiency(BE)of massive MIMO with perfect channel state information(CSI)by applying low resolution ADCs with Rician *** start our analysis by deriving the additive quantization noise model,which helps to understand the effects of ADC resolution on BE by keeping the power constraint at the receiver in *** also investigate deeply the effects of using higher bit rates and the number of BS antennas on bandwidth efficiency(BE)of the *** emphasize that good bandwidth efficiency can be achieved by even using low resolution ADC by using regularized zero-forcing(RZF)combining *** also provide a generic analysis of energy efficiency(EE)with different options of bits by calculating the energy efficiencies(EE)using the achievable *** emphasize that satisfactory BE can be achieved by even using low-resolution ADC/DAC in massive MIMO.

关键词： low-bit analog-digital converter massive(multiple-input-multiple-output)MIMO minimum mean square error(MMSE) regularized zero forcing zero forcing

来源：评论

学校读者我要写书评

暂无评论

Enhanced Textual Image Classification with Ensemble Learning of MobileNetV2 and ResNet50

Enhanced Textual Image Classification with Ensemble Learning...

引用

2024 International Conference on Intelligent Computing and Sustainable Innovations in technology, IC-SIT 2024

作者： Tripathy, Sasmita Pattanaik, Satya Ranjan Rabindranath, Mohanty Hitesh Dash, Ratnakar Centurion University of Technology and Management Department of Computer Science and Engineering Bhubaneswar India Gandhi Institute for Technology Department of Computer Science and Engineering Bhubaneswar India National Institute of Technology Rourkela Department of Computer Science and Engineering Rourkela India

ISBN: (纸本)9798350369175

Textual image classification is crucial in various applications, such as document digitization and automatic language identification. Although ensemble learning has been increasingly utilized to improve the accuracy of deep learning models, the optimal approaches for this task are still under-explored. This research evaluates the effectiveness of ensemble learning techniques for textual image classification, focusing on combining Convolutional Neural Networks (CNNs) like MobileNetV2 and ResNet50. The dataset used in this study includes images in multiple languages, adding complexity to the classification task. By leveraging ensemble learning, we enhanced the performance of these models, achieving an accuracy of 77%, surpassing the individual CNN models. These findings suggest that ensemble learning can significantly boost the effectiveness of deep learning models in textual image classification tasks. © 2024 IEEE.

关键词： Contrastive Learning

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：