检索结果-内蒙古大学图书馆

3rd International Conference on Machine Learning and computer Application, ICMLCA 2022

作者： Chen, Xiaoyang Li, Bokang Li, Weiheng School of Computer Science and Technology Tiangong University Tianjin China College of Information Science and Engineering Henan University of Technology Zhengzhou China College of Computer Science Chongqing University Chongqing China

ISBN: (纸本)9781510664814

FedAvg is a classic algorithm for Federated Learning. The MOON algorithm is based on the idea of comparative learning. The model in the previous round is a negative sample, and the global model is a positive sample, which improves the local training of FedAvg. However, in the study of MOON, it was found that since the MOON algorithm compares models rather than pictures, the number of negative samples is difficult to expand, and in traditional comparative learning, the total number of negative samples will significantly affect the effect of comparative learning, when the model in this round is far from the model in the previous round, it may approach the earlier model, thus reducing the efficiency of contrastive learning. So, this paper studied more algorithms related to contrastive learning and federated learning. Finally, a dictionary was built based on the MoCo algorithm to save the previously generated local models. The improved MOON algorithm was named FedCOMO. With the extension of negative samples, the training accuracy on the CIFAR-10 dataset can exceed MOON. © 2023 SPIE.

关键词： Moon

来源：评论

学校读者我要写书评

暂无评论

Detached and Interactive Multimodal Learning 24

Detached and Interactive Multimodal Learning

引用

32nd ACM International Conference on Multimedia, MM 2024

作者： Fan, Yunfeng Xu, Wenchao Wang, Haozhao Liu, Junhong Guo, Song Department of Computing The Hong Kong Polytechnic University Hong Kong Hong Kong School of Computer Science and Technology Huazhong University of Science and Technology Hubei China School of Software Beihang University Beijing China Department of Computer Science and Engineering Hong Kong University of Science and Technology Hong Kong Hong Kong

ISBN: (纸本)9798400706868

Recently, Multimodal Learning (MML) has gained significant interest as it compensates for single-modality limitations through comprehensive complementary information within multimodal data. However, traditional MML methods generally use the joint learning framework with a uniform learning objective that can lead to the modality competition issue, where feedback predominantly comes from certain modalities, limiting the full potential of others. In response to this challenge, this paper introduces DI-MML, a novel detached MML framework designed to learn complementary information across modalities under the premise of avoiding modality competition. Specifically, DI-MML addresses competition by separately training each modality encoder with isolated learning objectives. It further encourages cross-modal interaction via a shared classifier that defines a common feature space and employing a dimension-decoupled unidirectional contrastive (DUC) loss to facilitate modality-level knowledge transfer. Additionally, to account for varying reliability in sample pairs, we devise a certainty-aware logit weighting strategy to effectively leverage complementary information at the instance level during inference. Extensive experiments conducted on audio-visual, flow-image, and front-rear view datasets show the superior performance of our proposed method. The code is released at https://***/fanyunfeng-bit/DI-MML. © 2024 Owner/Author.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Digital Asset Management Tool for Film Animation and Game Development

引用

computer-Aided Design and Applications 2023年第S8期20卷 89-99页

作者： Cao, Ling Hao, Gang Liang, Yaner School of Information Engineering Guangzhou Vocational College of Technology & Business Guangdong Guangzhou511442 China School of computer Science Guangdong Polytechnic Normal University Guangdong Guangzhou510665 China

With the continuous development of China's economy and technology, China's market has also produced a large number of excellent film and television works with beautiful pictures. Once launched, it has been loved by consumers and promoted the further development of China's economy. Through research, it is found that this has a great relationship with the application of game engine to the design of film and television animation. This paper discusses the design and implementation of video animation of game engine. This paper uses the CAD game engine to create an animation example "Pipi's Creation" to illustrate the whole process of using the game engine to create film and television animation. In the process of creation, it focuses on describing the different processes of video animation produced by game engine and traditional animation, as well as the different links in the creation, and completes the creation of examples. Based on the above research, on the basis of 3D software CAD, a digital asset management system for computer-aided video animation and game development has been developed and implemented. © 2023 CAD Solutions, LLC.

关键词： computer aided design

来源：评论

学校读者我要写书评

暂无评论

Rolling Bearing Intelligent Faults Diagnosis Based on Deep Convolutional Neural Network

Rolling Bearing Intelligent Faults Diagnosis Based on Deep C...

引用

2022 International Conference on Signal Processing, computer Networks, and Communications, SPCNC 2022

作者： Wang, Wenjie Lv, Rongxin Ding, Haonan Li, Tengfeng Li, Tingyu Wang, Xin School of Computer Science Shenyang Aerospace University Shenyang China School of Innovation Shenyang Aerospace University Shenyang China School of Electrical Engineering Henan University of Technology Henan Zhengzhou China School of Resource and Security Engineering Henan Institute of Engineering Henan Zhengzhou China School of software Zhongyuan University of Technology Henan Zhengzhou China School of Information Science and Technology Zhengzhou Normal University Henan Zhengzhou China

ISBN: (纸本)9781510664616

In the processing, manufacturing, and production of modern fields, rolling bearings, the most basic module of most mechanical equipment, have a key role that cannot be ignored. This paper proposes three fault diagnosis model architectures for rolling bearings based on the deep convolutional neural network. Three models were tested on the industry-common Case Western Reserve university Dataset (CWRU). The original vibration signal acquisition and processing module mainly uses the vibration signal window translation method to complete the segmentation of overlapping signals and uses the Inception network to efficiently complete one-dimensional signal preprocessing. Finally, we use t-distributed stochastic neighbor embedding (t-SNE) to reduce the dimension and visualize the learned fault data distribution. © 2023 SPIE.

关键词： Roller bearings

来源：评论

学校读者我要写书评

暂无评论

GestureID:Gesture-Based User Authentication on Smart Devices Using Acoustic Sensing

引用

Sound & Vibration 2024年第1期58卷 151-169页

作者： Jizhao Liu Jiang Hui Zhaofa Wang School of Computer Science Zhongyuan University of TechnologyZhengzhou450007China Henan Development and Innovation Laboratory of Industrial Internet Security Big Data(Zhongyuan University of Technology) Zhengzhou450007China

User authentication on smart devices is crucial to protecting user privacy and device *** to the development of emerging attacks,existing physiological feature-based authentication methods,such as fingerprint,iris,and face recognition are vulnerable to forgery and *** this paper,GestureID,a system that utilizes acoustic sensing technology to distinguish hand features among users,is *** involves using a speaker to send acoustic signals and a microphone to receive the echoes affected by the reflection of the hand movements of the *** ensure system accuracy and effectively distinguish users’gestures,a second-order differential-based phase extraction method is *** method calculates the gradient of received signals to separate the effects of the user’s hand movements on the transmitted signal from the background ***,the secondorder differential phase and phase-dependent acceleration information are used as inputs to a Convolutional Neural Networks-Bidirectional Long Short-Term Memory(CNN-BiLSTM)model to model hand motion *** decrease the time it takes to collect data for new user registration,a transfer learning method is *** involves creating a user authentication model by utilizing a pre-trained gesture recognition *** a result,accurate user authentication can be achieved without requiring extensive amounts of training *** demonstrate that GestureID can achieve 97.8%gesture recognition accuracy and 96.3%user authentication accuracy.

关键词： Acoustic sensing hand gesture user authentication

来源：评论

学校读者我要写书评

暂无评论

Realistic Nighttime Haze Image Generation with Glow Effect 22

Realistic Nighttime Haze Image Generation with Glow Effect

引用

Proceedings of the 2022 11th International Conference on Networks, Communication and Computing

作者： Yading Zheng Aizhong Mi Yingxu Qiao Yijiang Wang School of Software Henan Polytechnic University China School of Computer Science and Technology Henan Polytechnic University China

ISBN: (纸本)9781450398039

Abstract: Haze rendering aims to generate realistic nighttime haze images from clear images. The results can be applied to various practical applications, such as nighttime image dehazing algorithms, game scene rendering, shooting filters, etc. We investigate two smaller but challenging problems in nighttime haze rendering, namely 1) how to accurately estimate the transmission map and air light from clear and haze images, respectively, in the absence of paired datasets? 2) How to render the characteristics of a realistic nighttime haze image: glow effect? For this purpose, we propose an unsupervised nighttime haze image rendering method called NHRM (Nighttime Haze Rendering Module). Precisely, the NHRM consists of three modules: 1) an illumination retention module based on Retinex theory. 2) a transmission map estimation network using multi-scale feature fusion and attention mechanism. 3) a glow generation module using tone mapping and convolution bloom technique. Compared with existing haze generation methods, NHRM can render realistic and controllable nighttime haze images by simulating the main features of nighttime haze scenes in an unsupervised manner. Numerous experimental results show that our method outperforms existing rendering methods with better visual effects and quantitative metrics.

关键词： Image Rendering Image Hazing Glow Effect Atmospheric Scattering Model

来源：评论

学校读者我要写书评

暂无评论

Gray level co-occurrence matrix and extreme learning machine for Alzheimer's disease diagnosis

引用

International Journal of Cognitive Computing in Engineering 2021年 2卷 116-129页

作者： Gao, Shuangshuang School of Computer Science and Technology Henan Polytechnic University Henan Jiaozuo 454000 China

Alzheimer's disease (AD) is a chronic neurodegenerative disease, which is one of the biggest challenges in geriatrics. The global incidence of Alzheimer's disease has been on the rise all the time. Since the cause of Alzheimer's disease is unknown and there is no cure at present, early diagnosis of AD is particularly important. computer researchers have developed various early detection methods based on machine learning and computer vision. In this study, we effectively combined the gray level co-occurrence matrix and the extreme learning machine to carry out automatic diagnosis of AD, used the gray level co-occurrence matrix to extract texture features, combined with the extreme learning machine to classify AD images, so as to obtain higher training speed and better generalization performance. The results of 10×10-fold cross validation experiments show that our method is superior to several advanced methods. Therefore, this method is effective in the detection of Alzheimer's disease, this method can effectively improve the classification accuracy, and also has good generalization. © 2021

关键词： Alzheimer's disease Detection Extreme learning machine Gray level co-occurrence matrix

来源：评论

学校读者我要写书评

暂无评论

Alignment efficient image-sentence retrieval considering transferable cross-modal representation learning

引用

Frontiers of computer science 2024年第1期18卷 77-91页

作者： Yang YANG Jinyi GUO Guangyu LI Lanyu LI Wenjie LI Jian YANG School of Computer Science and Engineering Nanjing University of Science and TechnologyNanjing 210094China Department of Computing Hong Kong Polytechnic UniversityHong Kong 100872China State Key Lab.for Novel Software Technology Nanjing UniversityNanjing 210094China 14th Research Institute of China Electronics Technology Group Corporation Nanjing 210094China guangyu.li2017@***

Traditional image-sentence cross-modal retrieval methods usually aim to learn consistent representations of heterogeneous modalities,thereby to search similar instances in one modality according to the query from another modality in *** basic assumption behind these methods is that parallel multi-modal data(i.e.,different modalities of the same example are aligned)can be obtained in *** other words,the image-sentence cross-modal retrieval task is a supervised task with the alignments as ***,in many real-world applications,it is difficult to realign a large amount of parallel data for new scenarios due to the substantial labor costs,leading the non-parallel multi-modal data and existing methods cannot be used *** the other hand,there actually exists auxiliary parallel multi-modal data with similar semantics,which can assist the non-parallel data to learn the consistent ***,in this paper,we aim at“Alignment Efficient Image-Sentence Retrieval”(AEIR),which recurs to the auxiliary parallel image-sentence data as the source domain data,and takes the non-parallel data as the target domain *** single-modal transfer learning,AEIR learns consistent image-sentence cross-modal representations of target domain by transferring the alignments of existing parallel ***,AEIR learns the image-sentence consistent representations in source domain with parallel data,while transferring the alignment knowledge across domains by jointly optimizing a novel designed cross-domain cross-modal metric learning based constraint with intra-modal domain adversarial ***,we can effectively learn the consistent representations for target domain considering both the structure and semantic ***,extensive experiments on different transfer scenarios validate that AEIR can achieve better retrieval results comparing with the baselines.

关键词： image-sentence retrieval transfer learning semantic transfer structure transfer

来源：评论

学校读者我要写书评

暂无评论

Improving Big Data Recommendation System Performance Using NLP Techniques With Multi-attributes

引用

Informatica (Slovenia) 2024年第5期48卷 63-70页

作者： Omar, Hoger K. Frikha, Mondher Jumaa, Alaa Khalil ENETCOM ATISP Research Lab University of Sfax Sfax Tunisia Department of Computer Science College of Computer Science and Information Technology University of Kirkuk Kirkuk Iraq Department of Electronics National School of Electronics and Telecommunications of Sfax University of Sfax Tunisia Database Technology Department Technical College of Informatics Sulaimani Polytechnic University Kurdistan Region Sulaimani46001 Iraq

Due to the wide availability of big data, institutions and companies are currently concentrating on developing highly effective recommender systems for their users. Traditional recommender systems use standard information such as user, item, and ratings. However, this data may not be sufficient for precise results. To enhance accuracy, it is recommended to include additional information such as textual data in the recommendation system. When dealing with large textual data, employing Natural Language Processing (NLP) techniques is essential for effective data analysis. Therefore, this work proposed a novel big data recommender system that enhances collaborative filtering (CF) results by leveraging NLP techniques and dealing with multiple attributes. The study constructs two big data recommendation system models using a machine learning algorithm. In both models, the Alternating Least Squares (ALS) algorithm within the Apache Spark big data tool has been used. The first model did not incorporate NLP techniques, while the second model considered the novel NLP techniques by taking into account the user's review comments. A dataset of more than 3 million ratings and reviews was gathered from the Amazon website with a size of 3.1 GB. The results showed significant improvement after incorporating the suggested NLP-based techniques with multiple attributes. © 2024 Slovene Society Informatika. All rights reserved.

关键词： Recommender systems

来源：评论

学校读者我要写书评

暂无评论

Typhoon Track Image Prediction Using a Comprehensive LSTM Attention Model in Deep Learning 4

Typhoon Track Image Prediction Using a Comprehensive LSTM At...

引用

4th International Conference on Neural Networks, Information and Communication Engineering, NNICE 2024

作者： Jiang, Yiping Yu, Yi Ma, Yuelong Zhuang, Jiahao Ye, Xianze Yuan, Yifeng Su, Guanpeng Guangdong University of Science and Technology School of Computer Science Guangdong Dongguan China University of Macau Faculty of Science and Technology Fst China South China Normal University School of Information and Optoelectronic Science and Engineering Guangdong Guangzhou China Macao Polytechnic University Faculty of Applied Sciences China Zhuhai College of Science and Technology School of Computer Guangdong Zhuhai China

ISBN: (纸本)9798350394375

Typhoons are exceptionally destructive tropical cyclones that possess the capacity to inflict significant harm upon society. For institutions engaged in risk assessment and disaster mitigation, accurately forecasting typhoon paths and trajectories is of paramount importance. In this study, we propose a Typhoon Track Image Prediction model based on a Comprehensive Long Short-Term Memory (LSTM) Attention Model in the field of Deep Learning. This model significantly improves the accuracy of typhoon track prediction. To assess its performance, we utilize typhoon data from the China Typhoon Network spanning from 1949 to 2022. The Comprehensive LSTM Attention Model is compared against other popular models such as Temporal Convolutional Networks (TCN), traditional LSTM, and Seq2Seq models for forecasting the typhoon track over a 12-hour period. The results demonstrate the superiority of our proposed model, showcasing its effectiveness in terms of superior accuracy in typhoon track prediction. © 2024 IEEE.

关键词： Long short-term memory

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：