检索结果-内蒙古大学图书馆

international conference on Fuzzy systems and Knowledge Discovery (FSKD)

作者： Luyao Chang Zhi Zhang College of Computer Science and Technology Wuhan University of Science and Technology Wuhan Hubei China

ISBN: (数字)9798350356328

ISBN: (纸本)9798350356335

To address the challenges of prolonged sampling time and the limitation to process specific image sizes in current image deraining research, we propose a novel single image deraining approach based on denoising diffusion implicit models. The method adopts a variant of the U-Net fully convolutional network architecture based on SR3, concatenating rainy images of arbi-trary sizes and noise images along the channel dimension to iteratively obtain high-resolution rain-free images. Specifically, we in-corporate residual blocks from BigGAN-deep to replace traditional residual blocks, thereby deepening the network to enhance model generalization. Additionally, we employ smooth noise esti-mation via image overlapping block processing to improve image fidelity. Introducing deterministic acceleration sampling, we expe-dite the sampling process by uniformly selecting time steps. Comprehensive testing on synthetic and real datasets demonstrates that the proposed method achieves deraining for images of any size, yielding significant improvements in peak signal-to-noise ra-tio and structural similarity. The generated images preserve more complete details, showing superior deraining efficacy.

关键词： Rain Computational modeling Noise reduction Noise Estimation Computer architecture Knowledge discovery image restoration Convolutional neural networks Fuzzy systems

来源：评论

学校读者我要写书评

暂无评论

Real-Time Object Detection in Overexposed images based on YOLOv5

Real-Time Object Detection in Overexposed Images Based on YO...

引用

Electronic technology, Communication and Information (ICETCI), IEEE international conference on

作者： Leyang Wang Jiaqi He Junhao Chen Wuhan University of Technology Wuhan China

ISBN: (数字)9798350361643

ISBN: (纸本)9798350361650

In the context of rapid advancements in the fields of the internet of Things (IoT) and artificial intelligence (AI), edge computing technology and its devices have undergone significant development and wide application, leading to the increasingly common deployment of lightweight neural networks on these devices. Despite this, research on real-time object detection tasks on edge computing devices under conditions of poor image quality remains relatively scarce. This is primarily because real-time object detection requires consideration of the computational power limitations of edge computing devices, making it challenging to satisfy both detection quality and speed simultaneously. In response to this challenge, this study proposes a YOLOv5 object detection method based on the Jetson Orin Nano edge computing device, aimed at addressing overexposure issues to ensure that detection accuracy is maintained while effectively solving overexposure correction problems.

关键词： YOLO Performance evaluation Accuracy image edge detection Computational modeling Lighting Real-time systems

来源：评论

学校读者我要写书评

暂无评论

Explaining Meta-Features Importance in Meta-Learning Through Shapley Values 25

Explaining Meta-Features Importance in Meta-Learning Through...

引用

25th international conference on Enterprise Information systems, ICEIS 2023

作者： Garouani, Moncef Ahmad, Adeel Bouneffa, Mourad Univ. Littoral Côte d'Opale LISIC Laboratoire d'Informatique Signal et Image de la Côte d'Opale France

ISBN: (纸本)9789897586484

Meta-learning, or the ability of a machine learning model to adapt and improve on a wide range of tasks, has gained significant attention in recent years. A crucial aspect of meta-learning is the use of meta-features, which are high-level characteristics of the data that can guide the learning process. However, it is a challenging task to determine the importance of different meta-features in a specific context. In this paper, we propose the use of Shapley values as a method for explaining the importance of meta-features in meta-learning process. Whereas, Shapley values is a well-established method in game theory. It has been used for fair distribution of payouts among a group of individuals, based on the separate contribution of meta-features to the overall payout. Recently, these have been also applied to machine learning to understand the contribution of different features in a model's prediction. We observe that a better understanding of meta-features, using the Shapely values, can be gained to evaluate their importance. In the context of meta-learning it may aid to improve the performance of the model. Our results demonstrate that Shapley values can provide insight into the relative importance of different meta-features and how they interact in the learning process. This can fairly optimize the meta-learning models, resulting in more accurate and effective predictions. Overall, this work conclude that Shapley values can be a useful tool in guiding the design of meta-features and these can be used to improve the performance of the meta-learning algorithms. Copyright © 2023 by SCITEPRESS - Science and technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

关键词： Game theory

来源：评论

学校读者我要写书评

暂无评论

Design of dynamic load balancing algorithm for heterogeneous clusters based on energy consumption

引用

ENERGY REPORTS 2021年 7卷 684-695页

作者： Ye, Lunqiang Southwest Minzu Univ Informat & Educ Technol Ctr Chengdu 610225 Sichuan Peoples R China

In order to reduce energy waste and improve the dynamic load balancing of heterogeneous clusters, sensor technology is introduced into the internet of Things technology to build a channel transmission model for heterogeneous clusters. Then use the dynamic weighting method to configure the output channel to complete the decomposition of the channel characteristics. In addition, based on the establishment of the channel fuzzy recombination structure model, the noise interference suppression method is used to suppress the multipath interference of heterogeneous cluster communication channels, and the equalization of the channel output is controlled by the baud interval equalization sampling method. At the same time, through the fuzzy balance configuration and spatial balance scheduling process, the dynamic load balancing processing of heterogeneous clusters is realized. The simulation results show that the method proposed in this paper has better balance for the load scheduling of heterogeneous clusters and reduces the bit error rate of the output signal. More importantly, the output balance and adaptive control capabilities of heterogeneous clusters have also been improved. (C) 2021 Published by Elsevier Ltd.

关键词： internet of Things Heterogeneous clusters Node sensors Dynamic load Balancing algorithms Data fusion algorithms

来源：评论

学校读者我要写书评

暂无评论

Photography Style Analysis using Convolutional Neural Networks

Photography Style Analysis using Convolutional Neural Networ...

引用

international IEEE conference on signal-image Technologies and internet-based System

作者： Michael Zouros Theodoros Giannakopoulos University of Piraeus Piraeus Greece Multimedia Analysis Group (MagCIL) Computational Intelligence Lab NCSR Demokritos Athens Greece

In this paper, we present a novel dataset of photographs annotated in terms of the respective image aesthetics. We also examine the ability of Convolutional Neural Networks (CNNs) to distinguish between the adopted photography style classes. In particular, we have defined five photography style classification tasks, related to the following aesthetic attributes: Color, Depth of Field (DoF), Palette, Composition and Type. We then followed an annotation procedure using on a set of 1832 photos selected from the Unsplash dataset. Multiple annotators have also been used, in order to measure inter-annotator agreement. As soon as the dataset was compiled, we trained and evaluated a Residual Neural Network (ResNet50). The experimental results prove that, despite the imbalanced dataset, our model was able to achieve acceptable classification results. The dataset is openly provided, along with the trained models and Python code to use them.

关键词： Photography Codes image color analysis Annotations Convolutional neural networks Task analysis Residual neural networks

来源：评论

学校读者我要写书评

暂无评论

FPGA Implementation of QC-LDPC Decoder with Rapid Cyclic Shifter for CV-QKD systems

FPGA Implementation of QC-LDPC Decoder with Rapid Cyclic Shi...

引用

Communication, image and signal Processing (CCISP), international conference on

作者： Zeyuan Jin Yinghua Xie Xue-Qin Jiang Genlong Chen School of Information Science and Technology Donghua University Shanghai China

ISBN: (数字)9798350356656

ISBN: (纸本)9798350356663

In continuous variable quantum key distribution (CV-QKD) systems, low-density parity check (LDPC) decoders have gained widespread adoption due to their robust error correction capabilities and good adaptability to parallel processing hardware designs. We introduce a syndrome-based normalized min-sum algorithm (S-NMSA) and implement the decoder on FPGAs for high throughput and minimal resource consumption. The decoder utilizes a semi-parallel architecture. The proposed rapid cyclic shifter (RCS) can complete the shift of any dimension matrix in one clock cycle and realizes a significant reduction in hardware resource consumption. The semi-parallel decoding structure attains a 68.5 Mbps processing throughput on the condition that the system clock of the quasi-cyclic (QC) LDPC decoder is 250 Mhz.

关键词： Phase change materials signal processing algorithms Throughput Parity check codes Hardware Decoding Quantum key distribution Resource management Field programmable gate arrays Clocks

来源：评论

学校读者我要写书评

暂无评论

Text-to-image Retrieval based on Zero-shot Transfer Learning with CLIP Model and Vector Database

Text-to-image Retrieval Based on Zero-shot Transfer Learning...

引用

international conference on Behavior, Economic and Social Computing (BESC)

作者： Junfeng Xie Junying Chen School of Software Engineering South China University of Technology Guangzhou China

ISBN: (数字)9798331531904

ISBN: (纸本)9798331531911

With the continuous increase in the number of internet users, the explosion of multimodal data on the web has led to a growing demand for image retrieval. Current text-to-image retrieval systems often rely on keyword matching, which frequently fails to accurately capture users' retrieval needs. This work proposes a text-to-image retrieval method based on zero-shot transfer learning with the CLIP model and vector database, designed to be applied in text-to-image retrieval systems. By leveraging the semantic alignment capability of the CLIP model for text and images, this study effectively addresses the semantic gap problem in text-to-image retrieval. Additionally, the method utilizes the Milvus database for efficient storage and retrieval of image feature vectors, thereby enhancing system stability. Experimental results on the Flickr30k dataset demonstrate that this method improves the accuracy of text-to-image retrieval, verifying its effectiveness. Furthermore, this work extends the proposed method by incorporating the iFLYTEK Spark Cognitive Model to support image retrieval using Chinese text descriptions, further broadening the system's applicability and reliability.

关键词： Accuracy Databases Transfer learning Semantics image retrieval Text to image Vectors Stability analysis Sparks Usability

来源：评论

学校读者我要写书评

暂无评论

Deep learning for image segmentation: Ultrasound image segmentation of thyroid nodules based on U-Net 6

Deep learning for image segmentation: Ultrasound image segme...

引用

6th international conference on Advances in image Processing, ICAIP 2022

作者： Liao, Yimou Lu, Tianxiu College of Mathematical and Statistics Sichuan University of Science and Engineering China China and Key Lab. of Higher Educ. of Sichuan Prov. for Entpr Informationalization and Internet of Things Key Laboratory of Higher Education of Sichuan Province China

ISBN: (纸本)9781450397155

Solving large system of tensor linear equations is a fundamental problem in mathematics. This paper proposes a sampling tensor greedy Kaczmarz method (tGK) to solve large-scale linear systems with a t-product structure by introducing an effective greedy criterion, which eliminates the entry with the largest residual in the submatrix system per iteration. Then a relaxed tensor greedy Kaczmarz method (tRGK (ω)) is obtained by introducing the relaxation parameter ω to tGK, which can effectively change the convergence rate. The linear convergence of the two methods is guaranteed when the tensor linear system is consistent. Several experiments show that the methods designed in this paper converge faster compared with tensor randomized Kaczmarz (tRK). Moreover, selecting appropriate parameters ω can improve the convergence rate of tGK. © 2022 ACM.

关键词： Linear systems

来源：评论

学校读者我要写书评

暂无评论

Automatic Music Classification Algorithm based on AI internet and image Processing technology

Automatic Music Classification Algorithm based on AI Interne...

引用

Distributed Computing and Electrical Circuits and Electronics (ICDCECE), IEEE international conference on

作者： Chen Wang School of Children’s Welfare Changchun Humanities and Sciences College Changchun China

ISBN: (数字)9798331533663

ISBN: (纸本)9798331533670

Automatic music classification algorithms have become a research hotspot in the field of music information retrieval. This study uses deep learning to denoise and normalize audio signals, and then uses Mel-frequency cepstral coefficients (MFCC) and image texture, color and other feature extraction techniques to establish rich feature representations for music works. On this basis, a music classification model is constructed using a long short-term memory network (LSTM) and the audio and image features are effectively fused through a late fusion strategy. The proposed algorithm shows excellent performance on multiple independent test sets, and the evaluation indicators such as classification accuracy, precision, recall rate, and F1 score all reach a high level, demonstrating the model's generalization ability for different music types.

关键词： Training Accuracy Computational modeling Scalability Feature extraction Classification algorithms Multiple signal classification Integrated circuit modeling Artificial intelligence Long short term memory

来源：评论

学校读者我要写书评

暂无评论

Semantic Communication based on Content Style Adaptation for image Transmission

Semantic Communication Based on Content Style Adaptation for...

引用

international conference on Wireless Communications and signal Processing (WCSP)

作者： Mengshu Song Nan Ma Yiming Liu Linzi Shen Chen Dong State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications Beijing China Department of Broadband Communication Peng Cheng Laboratory Shenzhen China

ISBN: (数字)9798350390643

ISBN: (纸本)9798350390650

The development of semantic communication has promoted the research on image transmission. Content and style are two crucial characteristics of image information. Existing studies have explored image information transmission and recovery methods in semantic communication systems, but do not consider both image style information and content information simultaneously. To address this issue, based on deep joint source-channel coding (JSCC) structure, a novel system named semantic communication based on content style adaption (CSASC) for image transmission is proposed in this paper. Specifically, a novel semantic information extraction method are designed to extract the semantic content information and semantic style information of images, meanwhile an image content-style adaptive reconstruction method is proposed by fusing semantic content and semantic style features to improve the image reconstruction accuracy. In addition, the impact of the weighting of semantic style and content information on image reconstruction quality is analyzed. Results demonstrate that the proposed method has better recovery results compared with existing image semantic communication methods.

关键词： Wireless communication image communication System performance Information processing Semantic communication Feature extraction Robustness Data mining image reconstruction signal to noise ratio

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：