检索结果-内蒙古大学图书馆

2023 IEEE World Conference on Communication and computing, WCONF 2023

作者： Balla, Sahithya Guru Prasad, M.S. Balla, Sahithi Madhu Krishna, K. Kiran, J. Ram, Mangey Dept. of Computer Science and Engg Dehradun India Reva University School of Computing and Information Technology Bengaluru India

ISBN: (纸本)9798350311204

This research paper presents a comparative analysis of emotion recognition using deep learning techniques. The aim of the study was to compare the performance of state-of-the-art deep learning models, namely Convolutional Neural Network (CNN), Mobilenet and Long Short-Term Memory (LSTM). The dataset used for our work is the popular FER - 2013 dataset, which consist of annotated tweets with emotion labels. The proposed work evaluates the models based on their precision, accuracy, recall, and F1-score. The results show that the CNN model, which was trained using the image dataset, outperforms the other models in terms of precision and F1-score. The study also analyzes the effect of various pre-processing techniques on the performance of the models. Overall, the study provides a comprehensive analysis of emotion recognition using deep learning models and highlights the strengths and weaknesses of different approaches. The results of this study are useful for researchers and practitioners working in the fields of natural language processing and emotion recognition. © 2023 IEEE.

关键词： Emotion Recognition

来源：评论

学校读者我要写书评

暂无评论

MuLAD: Multimodal Aggression Detection from Social Media Memes Exploiting Visual and Textual Features 27th

MuLAD: Multimodal Aggression Detection from Social Media Me...

引用

27th International Conference on Pattern Recognition, ICPR 2024

作者： Hasan, Md. Maruf Ahsan, Shawly Hoque, Mohammed Moshiul Dewan, M. Ali Akber Chittagong University of Engineering and Technology Chittagong4349 Bangladesh School of Computing and Information Systems Faculty of Science and Technology Athabasca University AthabascaABT9S 3A3 Canada

ISBN: (纸本)9783031781186

Aggression detection from memes is challenging due to their region-specific interpretation and multimodal nature. Detecting or classifying aggressive memes is complicated in low-resource languages (including Bengali) because benchmark datasets and primary language processing software are needed. This paper proposes an innovative meme classification technique that harnesses deep learning (DL) approaches to leverage memes’ visual and textual features in Bengali. Various DL frameworks, such as VGG16, VGG19, ResNet50, CNN, BiLSTM, and BiLSTM+CNN, extract visual and textual features from memes. A novel corpus named the Bengali Meme Dataset (AMemD) is also introduced, comprising a substantial amount of multimodal data, including text and image components. Experimental results on AMemD demonstrate the effectiveness of the proposed approach. The CNN combined with VGG16 obtained the highest f1-score of 0.738 among all multimodal techniques tested. This pioneering research offers valuable insights into the complex task of aggression detection from memes in Bengali and provides a foundation for future studies in this area. The dataset is available at https://***/Maruf089/Multimodal-Aggression-Detection. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

MEVTR: A Multilingual Model Enhanced With Visual Text Representations 30

MEVTR: A Multilingual Model Enhanced With Visual Text Repres...

引用

Joint 30th International Conference on Computational Linguistics and 14th International Conference on Language Resources and Evaluation, LREC-COLING 2024

作者： Wang, Xiaohua Fei, Wenlong Hu, Min Zhang, Qingyu Zhu, Aoqiang School of Computer Science and Information Engineering HeFei University of Technology China Anhui Province Key Laboratory of Affective Computing and Advanced Intelligent Machine China

ISBN: (纸本)9782493814104

The goal of multilingual modelling is to generate multilingual text representations for various downstream tasks in different languages. However, some state-of-the-art pre-trained multilingual models perform poorly on many low-resource languages due to the lack of representation space and model capacity. To alleviate this issue, we propose a Multilingual model Enhanced with Visual Text Representations (MEVTR), which complements textual representations and extends the multilingual representation space with visual text representations. First, the visual encoder focuses on the glyphs and structure of the text to obtain visual text representations, and the textual encoder obtains textual representations. Then, multilingual representations are enhanced by aligning and fusing visual text representations and textual representations. Moreover, we propose similarity constraint, a self-supervised task to prompt the visual encoder to focus on more additional information. Prefix alignment and multi-head bilinear module are designed to acquire an improved integration effect of visual text representations and textual representations. Experimental results indicate that MEVTR benefits from visual text representations and achieves significant performance gains in downstream tasks. In particular, in the zero-shot cross-lingual transfer task, MEVTR achieves results that outperform the state-of-the-art adapter-based framework without the target language adapter. © 2024 ELRA Language Resource Association: CC BY-NC 4.0.

关键词： Signal encoding

来源：评论

学校读者我要写书评

暂无评论

A New Method for Predefined-Time Synchronization Control of Memristive Cohen-Grossberg Neural Networks

A New Method for Predefined-Time Synchronization Control of ...

引用

2023 IEEE International Conference on Memristive computing and Applications, ICMCA 2023

作者： Liu, Aidi Yu, Wenyue Huang, Lili Guo, Jian Lei, Tengfei Zhao, Hui School of Computer and Information Engineering Qilu Institute of Technology Jinan China School of Computer and Information Engineering Qilu Institute of Technology China Collaborative Innovation Center of Memristive Computing Application School of Computer and Information Engineering Qilu Institute of Technology China School of Information Science and Engineering University of Jinan Shandong Provincial Key Laboratory of Network Based Intelligent Computing Jinan China

ISBN: (纸本)9798350316230

This article mainly conducts a series of research on the synchronization control of Cohen-Grossberg neural networks. Firstly, a memristive Cohen-Grossberg neural network model with time-varying delay was constructed, and an equivalent system was obtained by combining inverse function theory, differential inclusion, and set-valued mapping. Secondly, a new predefined-time stability theorem was proposed, and sufficient conditions were obtained to ensure that the drive-response system achieves predefined-time synchronization by designing appropriate control strategy. The simulation results show that the stability time of the error system can be adjusted by the user, verifying the correctness and of the theory. © 2023 IEEE.

关键词： Neural network models

来源：评论

学校读者我要写书评

暂无评论

Practical Byzantine fault-tolerant consensus algorithm based on k-Nearest Neighbor 7

Practical Byzantine fault-tolerant consensus algorithm based...

引用

7th IEEE information technology and Mechatronics Engineering Conference, ITOEC 2023

作者： Liu, Qingtai Yu, Jingang Shenyang Ligong University School of Information Science and Engineering Liaoning China Shenyang Institute of Computing Technology Liaoning China

ISBN: (纸本)9798350334197

In response to the traditional Byzantine fault-tolerant consensus algorithm, which suffers from performance bottleneck and improper selection of master nodes when the number of nodes increases, this paper proposes a k-Nearest Neighbor based PBFT consensus algorithm, which is called K-PBFT consensus algorithm in this paper. By introducing the k-Nearest Neighbor algorithm, the idea of discarding the master node is adopted to evaluate the behaviour of nodes in consensus, dynamically update the credit value of nodes, classify the node hierarchy and find neighboring nodes for all nodes to communicate so as to complete consensus. Through comparative simulation experimental analysis, the improved algorithm can reduce the time delay and improve the consensus efficiency. © 2023 IEEE.

关键词： Motion compensation

来源：评论

学校读者我要写书评

暂无评论

Comparison of CNN-based deep learning architectures for rice diseases classification

引用

Artificial Intelligence in Agriculture 2023年第3期9卷 22-35页

作者： Md Taimur Ahad Yan Li Bo Song Touhid Bhuiyan Department of Computer Science and Engineering Daffodil International UniversitySavarBangladesh School of Mathematics Physics and Computing University of Southern QueenslandToowoombaAustralia School of Engineering University of Southern Queensland ToowoombaAustralia Department of Computer Science and Engineering Faculty of Science and Information TechnologyDaffodil International UniversityBangladesh

Although convolutional neural network(CNN)paradigms have expanded to transfer learning and ensemble models from original individual CNN architectures,few studies have focused on the performance comparison of the applicability of these techniques in detecting and localizing rice ***,most CNN-based rice disease detection studies only considered a small number of diseases in their *** these shortcomings were addressed in this *** this study,a rice disease classification comparison of six CNN-based deep-learning architectures(DenseNet121,Inceptionv3,MobileNetV2,resNext101,Resnet152V,and Seresnext101)was conducted using a database of nine of the most epidemic rice diseases in *** addition,we applied a transfer learning approach to DenseNet121,MobileNetV2,Resnet152V,Seresnext101,and an ensemble model called DEX(Densenet121,EfficientNetB7,and Xception)to compare the six individual CNN networks,transfer learning,and ensemble *** results suggest that the ensemble framework provides the best accuracy of 98%,and transfer learning can increase the accuracy by 17%from the results obtained by Seresnext101 in detecting and localizing rice leaf *** high accuracy in detecting and categorisation rice leaf diseases using CNN suggests that the deep CNN model is promising in the plant disease detection domain and can significantly impact the detection of diseases in real-time agricultural *** research is significant for farmers in rice-growing countries,as like many other plant diseases,rice diseases require timely and early identification of infected diseases and this research develops a rice leaf detection system based on CNN that is expected to help farmers to make fast decisions to protect their agricultural yields and quality.

关键词： Deep learning Convolutional neural networks(CNNs) Transfer learning Plant leaf disease detection

来源：评论

学校读者我要写书评

暂无评论

A Knowledge-Based System to Predict Crime from Criminal Records in the Case of Hossana Police Commission 5

A Knowledge-Based System to Predict Crime from Criminal Reco...

引用

5th International Conference on information and Communication technology for Development for Africa, ICT4DA 2023

作者： Wubineh, Betelhem Zewdu Metekiya, Biniyam Yirsaw Wroclaw University of Science and Technology Faculty of Information and Communication Technology Wroclaw Poland School of Computing and Informatics Wachemo University Department of Information Technology Hossana Ethiopia College of Informatics Department of Information System University of Gondar Gondar Ethiopia

ISBN: (纸本)9798350328486

Crime is a complex social phenomenon, and its cost continues to rise over time. Police have collected a large amount of data on criminals every day for different crimes. However, extracting valuable information from these data takes time and effort for the law enforcement organization at hand. Therefore, effective crime prevention strategies and policies are necessary to prevent crime. Criminology is one of the most important areas for data mining. The purpose of this study is to extract useful crime patterns from past data and develop a knowledge-based system to prevent crime. Different criminals commit different crimes, but we do not know which criminal commits which crime. For this purpose, the 910 criminal records are used to train and test the model. To predict a crime, the classification technique and the J48 decision tree, random forest (RF), PART, and ensemble algorithms are used. As a result, the algorithm scored an accuracy of 75.38%, 80.4%,73.7%, and 81.3% in the J48, RF, PART, and ensemble methods, respectively, in a 10-fold cross-validation. Additionally, the offender's job is the determining factor that causes and predicts crimes. Finally, we integrate the rules extracted from a data mining task into a knowledge base system using PROLOG, which is used to make decisions. © 2023 IEEE.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

NeuVSA: A Unified and Efficient Accelerator for Neural Vector Search 31

NeuVSA: A Unified and Efficient Accelerator for Neural Vecto...

引用

31st IEEE International Symposium on High Performance Computer Architecture, HPCA 2025

作者： Yuan, Ziming Dai, Lei Li, Wen Zhang, Jie Liang, Shengwen Wang, Ying Liu, Cheng Li, Huawei Li, Xiaowei Guo, Jiafeng Wang, Peng Chen, Renhai Zhang, Gong Shanxi University School of Computer and Information Technology China Peking University School of Computer Science China Institute of Computing Technology CAS State Key Lab of Processors China University of Chinese Academy of Sciences China Zhongguancun National Laboratory Beijing China Institute of Computing Technology CAS Key Lab of Network Data Science and Technology China Huawei Technologies Co. Ltd. China

ISBN: (纸本)9798331506476

Neural Vector Search (NVS) has exhibited superior search quality over traditional key-based strategies for information retrieval tasks. An effective NVS architecture requires high recall, low latency, and high throughput to enhance user experience and cost-efficiency. However, implementing NVS on existing neural network accelerators and vector search accelerators is sub-optimal due to the separation between the embedding stage and vector search stage at both algorithm and architecture levels. Fortunately, we unveil that Product Quantization (PQ) opens up an opportunity to break separation. However, existing PQ algorithms and accelerators still focus on either the embedding stage or the vector search stage, rather than both simultaneously. Simply combining existing solutions still follows the beaten track of separation and suffers from insufficient parallelization, frequent data access conflicts, and the absence of scheduling, thus failing to reach optimal recall, latency, and throughput. To this end, we propose a unified and efficient NVS accelerator dubbed NeuVSA based on algorithm and architecture co-design philosophy. Specifically, on the algorithm level, we propose a learned PQ-based unified NVS algorithm that consolidates two separate stages into the same computing and memory access paradigm. It integrates an end-to-end joint training strategy to learn the optimal codebook and index for enhanced recall and reduced PQ complexity, thus achieving smoother acceleration. On the architecture level, we customize a homogeneous NVS accelerator based on the unified NVS algorithm. Each sub-accelerator is optimized to exploit all parallelism exposed by unified NVS, incorporating a structured index assignment strategy and an elastic on-chip buffer to alleviate buffer conflicts for reduced latency. All sub-accelerators are coordinated using a hardware-aware scheduling strategy for boosted throughput. Experimental results show that the joint training strategy improves recall

关键词： Network embeddings

来源：评论

学校读者我要写书评

暂无评论

Applications of electromagnetic metamaterials in reducing the size of antennas 9

Applications of electromagnetic metamaterials in reducing th...

引用

9th International Conference on Energy Materials and Electrical Engineering, ICEMEE 2023

作者： Pang, Xudong Wu, Liping Yuan, Shanshan Zhu, Shouzheng Department of Computing Center Shanghai Publishing and Printing College Shanghai China School of Information Science and Technology East China Normal University Shanghai China

ISBN: (纸本)9781510672765

Metamaterials are artificially designed materials with special electromagnetic properties that can manipulate the propagation paths of electromagnetic waves. The fabrication of metamaterials involves various physical processes, such as the adjustment of microstructure, lattice structure, and even adjustments at molecular level. By precisely controlling these processes and parameters, specific electromagnetic properties of metamaterials can be designed. The electromagnetic wave concentrator discussed in this paper is an electromagnetic metamaterial device, which can be used to reduce the size of the antenna, where the ability of miniaturization depends on the magnification of the device. The use of metamaterials can improve the directivity of antennas with the same size, which has a practical value for applications. © COPYRIGHT SPIE. Downloading of the abstract is permitted for personal use only.

关键词： antenna size Metamaterials transformation electromagnetics wave concentrator

来源：评论

学校读者我要写书评

暂无评论

Android Malware Detection System using Machine Learning 2024

Android Malware Detection System using Machine Learning

引用

16th International Conference on Contemporary computing, IC3 2024

作者： Kaur, Amanpreet Lal, Sangeeta Goel, Shruti Pandey, Mrinal Agarwal, Astha Department of Computer Science and Engineering & IT Jaypee Institute of Information Technology India School of computing and mathematics Keele University Newcastle-under-Lyme United Kingdom

ISBN: (纸本)9798400709722

Detecting Android malware is imperative for safeguarding user privacy, securing data, and preserving device performance. Consequently, numerous studies have underscored the complexities associated with Android malware detection, prompting a multidimensional approach to tackle these challenges effectively. This research leverages machine learning techniques, emphasizing feature extraction, classification algorithms, and both supervised and unsupervised learning methodologies. The exploration begins with in-depth Exploratory Data Analysis (EDA) to gain insights into the dataset, paving the way for informed decision-making. Principal Component Analysis (PCA) is employed for dimensionality reduction, a pivotal step in handling the multivariate nature of the data. The integration of API calls, clustering, and anomaly detection further enriches the model's capability to discern between benign and malicious applications. Crucially, the study delves into the intricacies of sampling, evaluation, and the Confusion Matrix to quantify the model's performance accurately. The utilization of diverse classification algorithms, including Support Vector Machines (SVM), Multi-Layer Perceptrons (MLP), Random Forest, GaussianNB, Decision Tree, and Logistic Regression, underscores the comprehensive nature of the approach. These algorithms collectively contribute to a robust and versatile Android malware detection model capable of adapting to varying threat scenarios. The dataset employed for training and evaluation is sourced from Kaggle, encompassing 29,999 Android applications categorized as benign or malicious based on permissions sought. Current detection methods, deemed resource-intensive and exhaustive, face the challenge of keeping pace with the relentless evolution of new malware strains. This research seeks to address this gap by proposing a sophisticated, machine learning-driven model that not only enhances accuracy but also demonstrates efficiency and adaptability in the face

关键词： Android malware

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：