检索结果-内蒙古大学图书馆

17th ACM conference on Recommender systems (RecSys)

作者： xi, Xumei Zhao, Yuke Liu, Quan Ouyang, Liwen Wu, Yang Cornell Univ Ithaca NY 14853 USA Bloomberg New York NY USA

ISBN: (纸本)9798400702419

We consider the problem of sequential recommendation, where the current recommendation is made based on past interactions. This recommendation task requires efficient processing of the sequential data and aims to provide recommendations that maximize the long-term reward. To this end, we train a farsighted recommender by using an offline RL algorithm with the policy network in our model architecture that has been initialized from a pre-trained transformer model. The pre-trained model leverages the superb ability of the transformer to process sequential information. Compared to prior works that rely on online interaction via simulation, we focus on implementing a fully offline RL framework that is able to converge in a fast and stable way. Through extensive experiments on public datasets, we show that our method is robust across various recommendation regimes, including e-commerce and movie suggestions. Compared to state-of-the-art supervised learning algorithms, our algorithm yields recommendations of higher quality, demonstrating the clear advantage of combining RL and transformers.

关键词： recommender systems reinforcement learning transformers

来源：评论

学校读者我要写书评

暂无评论

Optimal Transport-Guided Conditional Score-Based Diffusion Model 37

Optimal Transport-Guided Conditional Score-Based Diffusion M...

引用

37th conference on Neural Information processing systems (NeurIPS)

作者： Gu, xiang Yang, Liwei Sun, Jian Xu, Zongben Xi An Jiao Tong Univ Sch Math & Stat Xian Peoples R China Pazhou Lab Huangpu Guangzhou Peoples R China Peng Cheng Lab Shenzhen Peoples R China

ISBN: (纸本)9781713899921

Conditional score-based diffusion model (SBDM) is for conditional generation of target data with paired data as condition, and has achieved great success in image translation. However, it requires the paired data as condition, and there would be insufficient paired data provided in real-world applications. To tackle the applications with partially paired or even unpaired dataset, we propose a novel Optimal Transport-guided Conditional Score-based diffusion model (OTCS) in this paper. We build the coupling relationship for the unpaired or partially paired dataset based on L2-regularized unsupervised or semi-supervised optimal transport, respectively. Based on the coupling relationship, we develop the objective for training the conditional score-based model for unpaired or partially paired settings, which is based on a reformulation and generalization of the conditional SBDM for paired setting. With the estimated coupling relationship, we effectively train the conditional score-based model by designing a "resampling-by-compatibility" strategy to choose the sampled data with high compatibility as guidance. Extensive experiments on unpaired super-resolution and semi-paired image-to-image translation demonstrated the effectiveness of the proposed OTCS model. From the viewpoint of optimal transport, OTCS provides an approach to transport data across distributions, which is a challenge for OT on large-scale datasets. We theoretically prove that OTCS realizes the data transport in OT with a theoretical bound. Code is available at https://***/XJTU- XGU/OTCS.

关键词： Large datasets

来源：评论

学校读者我要写书评

暂无评论

Brain image Denoising and Segmentation Using Advanced Mixed Thresholding and Edge Preservation Analysis Technique

Brain Image Denoising and Segmentation Using Advanced Mixed ...

引用

2024 International conference on Knowledge Engineering and Communication systems, ICKECS 2024

作者： Raman, Ramakrishnan Kumar, Vikram Pillai, Biju G. Rabadiya, Dhaval Patre, Smruti Meenakshi, R. India Srv Media Pvt. Ltd India India Parul Institute of Technology Parul University Computer Science and Engineering Faculty of Engineering and Technology India Symbiosis Institute of Business Management Nagpur India Chennai Institute of Technology Computer Science and Engineering India

ISBN: (纸本)9798350359688

In the realm of medical imaging, particularly brain imaging, the clarity and accuracy of images are paramount for effective diagnosis and treatment planning. However, the inherent presence of noise in imaging modalities such as MRI often degrades the quality of these crucial images, complicating the segmentation process and potentially leading to inaccurate diagnoses. Addressing this challenge, our research introduces an innovative approach that synergizes advanced mixed thresholding with edge preservation analysis to significantly enhance brain image denoising and segmentation. This technique leverages a novel mixed thresholding method that dynamically adjusts to the specific noise characteristics of brain images, effectively reducing noise while retaining critical image details. Concurrently, our edge preservation analysis ensures that essential structural information is maintained, preventing the loss of vital features during the denoising process. The combination of these methods provides a robust solution to the persistent issue of noise in brain imaging, promising to improve the accuracy of segmentation and, by extension, the reliability of medical diagnoses derived from these images. By integrating advanced algorithms and mathematical models, our approach not only outperforms existing denoising and segmentation techniques but also establishes a new benchmark for image processing in medical diagnostics. © 2024 IEEE.

关键词： image denoising

来源：评论

学校读者我要写书评

暂无评论

Anomaly Detection in Load Forecasting for Electric Vehicles Using image processing Techniques 24

Anomaly Detection in Load Forecasting for Electric Vehicles ...

引用

24th EEEIC International conference on Environment and Electrical Engineering and 8th I and CPS Industrial and Commercial Power systems Europe, EEEIC/I and CPS Europe 2024

作者： Marandi, Saba Moradzadeh, Arash Moayyed, Hamed Assi, Chadi Ghafouri, Mohsen Vale, Zita Concordia University Ciise Department Montreal Canada Qatar University Electrical Engineering Department Doha Qatar Polytechnic of Porto GECAD-LASI Department Porto Portugal

ISBN: (纸本)9798350355185

The incorporation of electric vehicles (EVs) into the power grid presents significant challenges, particularly in meeting the electricity requirements of EVs through effective distribution strategies, efficient energy allocation, and optimal charging station placement. Accurate forecasting of load demand is critical for both the operational and planning aspects of energy systems, particularly due to the influence of EV travel behavior. This study introduces a hybrid approach to EV load forecasting using deep learning models to tackle this challenge. The proposed CLSTM model integrates convolutional neural network (CNN) and long short-term memory (LSTM) algorithms. Furthermore, the study introduces and applies an anomaly detection model based on image processing. The forecasting process initially uses clean data and then advances to a scenario involving a simulated cyber-attack by manipulating certain input data, which is referred to as a false data injection attack. After analyzing the data using various evaluation metrics and confirming anomalies, the image processing model identifies these anomalies through data visualization. In conclusion, a thorough evaluation of the results under both clean and compromised forecasting conditions unequivocally demonstrates that the CLSTM model not only surpasses the performance of traditional models but also exhibits greater robustness and accuracy. This superior performance highlights the CLSTM model's potential as a more reliable and effective solution for forecasting in scenarios impacted by data integrity issues. © 2024 IEEE.

关键词： Vehicle-to-grid

来源：评论

学校读者我要写书评

暂无评论

International conference on Integrated Intelligence and Communication systems, ICIICS 2023

International Conference on Integrated Intelligence and Comm...

引用

2023 International conference on Integrated Intelligence and Communication systems, ICIICS 2023

ISBN: (纸本)9798350315455

The proceedings contain 286 papers. The topics discussed include: utilizing machine learning to analyze and implement task scheduling algorithms in cloud;real time fire detection and alert triggering system;a dynamic model for blended learning of university English based on K-Means clustering algorithm;real time eye-tracking mouse control using recurrent neural network;collaborative virtual storage technology based on mobile cloud computing;enhanced intrusion detection system using Pearson correlation based long short-term memory;construction of group algorithm model under green innovation capability technology;numerical simulation and analysis of integrated mobile medical building based on finite element method;the implementation and optimization of high-performance image processing algorithms based on artificial intelligence;and research on photographic image classification based on multi-model fusion and data augmentation.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Wavelet based Convolutional Autoencoder for Medical image Compression

Wavelet based Convolutional Autoencoder for Medical Image Co...

引用

2023 International conference on Intelligent systems for Communication, IoT and Security, ICISCoIS 2023

作者： Venugopal, Muthalaguraja Palanisamy, Kalavathi Department of Computer Science and Applications Gandhigram Tamil Nadu India

ISBN: (纸本)9798350335835

Deep learning approaches have seen tremendous progress lately, particularly in image processing. In image compression, the deep autoencoders are predominantly showing promising results. Autoencoders have demonstrated exceptional performance in image compression tasks when compared to conventional lossy algorithms used for compression, in both the quantitative and qualitative aspects, despite the fact that they are lossy image compression approaches. This paper, puts forth a wavelet-based autoencoder-style compression approach for medical image compression. The proposed architecture is composed of four modules namely Pre-processing, Encoder, Decoder and Enhancement module. In order to achieve remarkable compression performance on the medical images, this method utilizes the advantage of wavelet decomposition technique to extract finer details from the low and high frequency components of the medical image. The experimental findings suggest that our approach achieves notable image compression performance. It outperforms the established JPEG in terms of the objective and subjective metrics PSNR and SSIM, and it also yields results that are comparable to those of JPEG2000 and BPG. © 2023 IEEE.

关键词： image compression

来源：评论

学校读者我要写书评

暂无评论

Plant Leaf Disease Identification Method Using Computer Vision and Machine Learning algorithms

Plant Leaf Disease Identification Method Using Computer Visi...

引用

2023 International conference on Integrated Intelligence and Communication systems, ICIICS 2023

作者： Sohan, Ismail Hossain Rahman, Saidur Chhabra, Gunjan Kaushik, Keshav Haque, Rezaul Dept. of Cse Dhaka Bangladesh Graphic Era Hill University Dept. of Cse Dehradun India University of Petroleum and Energy Studies School of Computer Science Dehradun India East West University Dept. of Cse Dhaka Bangladesh

ISBN: (纸本)9798350315455

This paper proposes an image processing-based method for detecting plant diseases. The system's three key phases are acquisition, pre-processing, and disease identification. Pictures of plant leaves are taken during the image acquisition stage using a digital camera. The image preprocessing stage involves enhancing and pre-processing the collected images to eliminate noise and extraneous data. The pre-processed photos are finally sorted into healthy or unhealthy, utilizing machine learning techniques in the illness identification step. This paper will represent how we can detect plant leaf diseases using a specific software system. It will also find the percentage of accuracy and error of disease detection based on the number of detected images. The leaf disease finding procedure information will also be explained here step by step. According to experimental data, the proposed approach has a high level of accuracy in detecting plant illnesses. Farmers can utilize the system as a valuable tool to detect plant illnesses early, minimizing yield losses and enhancing crop management techniques. © 2023 IEEE.

关键词： Disease detection Feature extraction image classification image processing Machine Learning

来源：评论

学校读者我要写书评

暂无评论

Ultra-high spatio-temporal resolution imaging with parallel acquisition-readout structured illumination microscopy(PAR-SIM)

引用

Light(Science & Applications) 2024年第6期13卷 1140-1158页

作者： xinzhu Xu Wenyi Wang Liang Qiao Yunzhe Fu xichuan Ge Kun Zhao Karl Zhanghao Meiling Guan xin Chen Meiqi Li Dayong Jin Peng xi Department of Biomedical Engineering College of Future TechnologyPeking UniversityBeijing100871China Wallace H.Coulter Dept.of Biomedical Engineering Georgia Institute of Technology and Emory UniversityAtlanta30332GAUSA National Biomedical Imaging Center Peking UniversityBeijing100871China Department of Biomedical Engineering College of EngineeringSouthern University of Science and TechnologyShenzhen518055GuangdongChina Airy Technologies Co. Ltd.Beijing100086China Eastern Institute for Advanced Study Eastern Institute of TechnologyNingboZhejiang315200China School of Life Science Peking UniversityBeijing100871China Institute for Biomedical Materials and Devices(IBMD) Faculty of ScienceUniversity of Technology SydneySydneyNSW2007Australia

Structured illumination microscopy(SIM)has emerged as a promising super-resolution fluorescence imaging technique,offering diverse configurations and computational strategies to mitigate phototoxicity during real-time imaging of biological *** efforts to enhance system frame rates have concentrated on processing algorithms,like rolling reconstruction or reduced frame reconstruction,or on investments in costly sCMOS cameras with accelerated row readout *** this article,we introduce an approach to elevate SIM frame rates and region of interest(ROI)coverage at the hardware level,without necessitating an upsurge in camera expenses or intricate ***,parallel acquisition-readout SIM(PAR-SIM)achieves the highest imaging speed for fluorescence imaging at currently available detector *** using the full frame-width of the detector through synchronizing the pattern generation and image exposure-readout process,we have achieved a fundamentally stupendous information spatial-temporal flux of 132.9 MPixels·s^(−1),9.6-fold that of the latest techniques,with the lowest SNR of−2.11 dB and 100 nm ***-SIM demonstrates its proficiency in successfully reconstructing diverse cellular organelles in dual excitations,even under conditions of low signal due to ultra-short exposure ***,mitochondrial dynamic tubulation and ongoing membrane fusion processes have been captured in live COS-7 cell,recorded with PAR-SIM at an impressive 408 *** posit that this novel parallel exposure-readout mode not only augments SIM pattern modulation for superior frame rates but also holds the potential to benefit other complex imaging systems with a strategic controlling approach.

关键词： hardware illumination resolution

来源：评论

学校读者我要写书评

暂无评论

Federated heuristic optimization based on fuzzy clustering and red fox optimization algorithm

Federated heuristic optimization based on fuzzy clustering a...

引用

IEEE International conference on Fuzzy systems (FUZZ-IEEE)

作者： Polap, Dawid Prokop, Katarzyna Srivastava, Gautam Silesian Tech Univ Fac Appl Math Kaszubska 23 PL-44100 Gliwice Poland Brandon Univ Dept Math & Comp Sci Brandon MB R7A 6A9 Canada

ISBN: (纸本)9798350332285

Heuristic algorithms are dependent on many coefficients like the number of iterations or individuals. However, quite often these algorithms move individuals toward the best in the population. Based on this observation, we propose the idea of federated heuristics. The proposed idea is to initially distribute individuals into certain intervals. Then, it performs a specified number of iterations of the algorithm to identify the potentially best intervals. Sorted intervals (in relation to the best-adapted individual) make it possible to separate the appropriate size of the population in each of them. Moreover, these clusters are merged by a fuzzy algorithm due to a decrease in their numbers. The more significant the interval, the greater the number of individuals and iterations allocated in these areas. As a consequence, several instances of the selected heuristic algorithm are triggered, which can divide the best individual. The proposed technique was described using the red fox algorithm and tested at a classic set of functions with different parameters of the used heuristic.

关键词： fuzzy neural network heuristic image processing

来源：评论

学校读者我要写书评

暂无评论

Cones 2: Customizable image Synthesis with Multiple Subjects 37

Cones 2: Customizable Image Synthesis with Multiple Subjects

引用

37th conference on Neural Information processing systems (NeurIPS)

作者： Liu, Zhiheng Zhang, Yifei Shen, Yujun Zheng, Kecheng Zhu, Kai Feng, Ruili Liu, Yu Zhao, Deli Zhou, Jingren Cao, Yang USTC Hefei Peoples R China SJTU Shanghai Peoples R China Ant Grp Hangzhou Peoples R China Alibaba Grp Hangzhou Peoples R China Alibaba DAMO Acad Hangzhou Peoples R China

ISBN: (纸本)9781713899921

Synthesizing images with user-specified subjects has received growing attention due to its practical applications. Despite the recent success in single subject customization, existing algorithms suffer from high training cost and low success rate along with increased number of subjects. Towards controllable image synthesis with multiple subjects as the constraints, this work studies how to efficiently represent a particular subject as well as how to appropriately compose different subjects. We find that the text embedding regarding the subject token already serves as a simple yet effective representation that supports arbitrary combinations without any model tuning. Through learning a residual on top of the base embedding, we manage to robustly shift the raw subject to the customized subject given various text conditions. We then propose to employ layout, a very abstract and easy-to-obtain prior, as the spatial guidance for subject arrangement. By rectifying the activations in the cross-attention map, the layout appoints and separates the location of different subjects in the image, significantly alleviating the interference across them. Both qualitative and quantitative experimental results demonstrate our superiority over state-of-the-art alternatives under a variety of settings for multi-subject customization. Project page can be found here.

关键词： Embeddings

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：