检索结果-内蒙古大学图书馆

A CVAE-within-Gibbs sampler for Bayesian linear inverse problems with hyperparameters

COMPUTATIONAL & APPLIED MATHEMATICS 2023年第3期42卷 1-23页

作者： Yang, Jingya Niu, Yuanling Zhou, Qingping Cent South Univ Sch Math & Stat HNP LAMA Changsha 410083 Hunan Peoples R China

We propose a conditional variational auto-encoder within Gibbs sampling (CVAE-within-Gibbs) for Bayesian linear inverse problems where the prior or the likelihood function depends on ambiguous hyperparameters. The method builds on ideas from classical sampling theory and recent advances in deep generative models to approximate complicated probability distributions. Specifically, we use a CVAE model which is trained with a large amount of data to learn the conditional density of hyperparameters in the original Gibbs sampler. The learned property of the conditional posterior provides more flexibility than classical Gibbs sampling because it avoids manually or experimentally determining the hyperpriors and their hyperparameters. We demonstrate the performance of the proposed method for three linear inverse problems, i.e., image deblurring, signal denoising, and boundary heat flux identification in a heat conduction problem.

关键词： Inverse problems Bayesian inference Gibbs sampling conditional variational auto-encoder Hyperparameters

来源：评论

学校读者我要写书评

暂无评论

A Fast Inverse Design Method for Reconfigurable Intelligent Metasurface Based on Generative Model 11

A Fast Inverse Design Method for Reconfigurable Intelligent ...

引用

11th IEEE MTT-S International Wireless Symposium (IWS) part of China Microwave Week

作者： Zhao, YiHan Lu, YingJuan Dai, JunYan Luo, ZhangJie Cheng, Qiang Southeast Univ State Key Lab Millimeter Waves Nanjing 210018 Peoples R China

ISBN: (纸本)9798350389999;9798350390001

Traditional design of reconfigurable intelligent metasurface (RIS) with specific scattering properties is a time-consuming process that relies on the design experience. In this article, we propose an inverse design method for RIS based on generative model to solve the non-uniqueness problem that one electromagnetic (EM) response corresponds to multiple unit structures. This method includes three parts: a forward prediction model, a conditional variational auto-encoder (CVAE) model and a Gumbel-Sinkhorn operator model. The forward prediction model is trained to quickly obtain the corresponding EM responses of RIS patterns. The CVAE model first encodes the RIS pattern and EM responses together into latent space and flexibly generates a series of corresponding candidate RIS patterns based on the target EM response. The Gumbel-Sinkhorn operator is used to reduce the discrete mapping error of these candidate RIS patterns. Then we feed these candidate patterns generated by the CVAE model and optimized by the Gumbel-Sinkhorn operator into the trained forward prediction model to obtain their corresponding EM responses. By selecting the predicted responses that fulfill the target response, the non-uniqueness problem is solved, which verifies the practicality and accuracy of our proposed method.

关键词： conditional variational auto-encoder Gumbel-Sinkhorn Operator Reconfigurable Intelligent Metasurface Fast Inverse Design

来源：评论

学校读者我要写书评

暂无评论

An ensemble active learning for a fluidized bed granulation in the pharmaceutical industry

引用

JOURNAL OF PROCESS CONTROL 2022年 118卷 16-25页

作者： Chen, Zhongxin Tang, Yongwei Gao, Zenglin Zhou, Jun Huang, Panling Shandong Univ Sch Mech Engn Jinan 250061 Peoples R China Minist Educ Key Lab High Efficiency & Clean Mech Manufacture Jinan 250061 Peoples R China Qilu Univ Technol Shandong Acad Sci Shandong Key Lab Comp Networks Jinan 250014 Peoples R China

Active learning (AL), allowing to query the labels by an oracle, can dwindle the generalization error from a few labeled samples, thus well-motivating in the scenario where there are fewer labeled samples and abundant unlabeled samples. However, the query samples cannot be exploited efficiently in the existing methods based on AL for regression tasks, and AL is typically applied for classification problems. In this paper, an active learning framework (CALF) is proposed to improve the efficiency with fewer query samples, aiming at the moisture content prediction in fluidized bed granulation. The proposed method, based on conditional variational auto-encoder (CVAE) and selective ensemble algorithm, can efficiently incorporate the information in query samples into the labeled sample space. The game of CVAE and selective ensemble algorithm improves the performance of the framework effectively, and its effectiveness is verified by the results obtained from the large batch of fluidized bed granulating experiments. (C) 2022 Elsevier Ltd. All rights reserved.

关键词： Active learning Pharmaceutical technology Selective ensemble conditional variational auto-encoder

来源：评论

学校读者我要写书评

暂无评论

Data-driven affective product design using complete three-dimensional surface data

引用

JOURNAL OF INTELLIGENT & FUZZY SYSTEMS 2022年第6期42卷 5437-5455页

作者： Wang, Zeng Liu, Weidong Yang, Minglang Nanchang Univ Adv Mfg Sch Nanchang Jiangxi Peoples R China Nanchang Univ Architecture & Design Coll Nanchang Jiangxi Peoples R China

As the main part of design display and evaluation, product three-dimensional (3D) form is the core object in affective product design. However, previous research has not yet addressed the development of technical models and method involving complete 3D surface data, and thus cannot guarantee the quality of affective product design. By using the techniques of triangular mesh model, spherical harmonic and conditional variational auto-encoder, this paper proposes a data-driven affective product design method composed of several technical models using complete 3D surface data. These models include: mathematical model for quantifying 3D form, recognition model for recognizing customer's affective responses, and generative model for generating new 3D forms. For affective product design, the mathematical model achieves the acquisition and processing of complete 3D surface data, the recognition model improves the objectivity and accuracy of recognition by integrating the 3D form data into the calculation process of emotion recognition, and the generative model realizes the automatic generation of new 3D forms in response to emotional data based on the recognition results. Each model provides technical support for realizing the acquisition, processing and generation of complete 3D surface data of product form, and ensures the systematicness and completeness of the proposed method for the affective product design involving 3D form innovation. The feasibility of the method is verified by an example of car design, and the results show that it is an effective affective product design method involving 3D form innovation.

关键词： Affective product form design complete three-dimensional surface data emotion recognition spherical harmonic conditional variational auto-encoder

来源：评论

学校读者我要写书评

暂无评论

variational Abnormal Behavior Detection With Motion Consistency

引用

IEEE TRANSACTIONS ON IMAGE PROCESSING 2022年 31卷 275-286页

作者： Li, Jing Huang, Qingwang Du, Yingjun Zhen, Xiantong Chen, Shengyong Shao, Ling Tianjin Univ Technol Sch Comp Sci & Engn Tianjin 300384 Peoples R China Nanchang Univ Sch Informat Engn Nanchang 330031 Jiangxi Peoples R China Univ Amsterdam AIM Lab NL-1098 XH Amsterdam Netherlands Incept Inst Artificial Intelligence Abu Dhabi U Arab Emirates

Abnormal crowd behavior detection has recently attracted increasing attention due to its wide applications in computer vision research areas. However, it is still an extremely challenging task due to the great variability of abnormal behavior coupled with huge ambiguity and uncertainty of video contents. To tackle these challenges, we propose a new probabilistic framework named variational abnormal behavior detection (VABD), which can detect abnormal crowd behavior in video sequences. We make three major contributions: (1) We develop a new probabilistic latent variable model that combines the strengths of the U-Net and conditional variational auto-encoder, which also are the backbone of our model;(2) We propose a motion loss based on an optical flow network to impose the motion consistency of generated video frames and input video frames;(3) We embed a Wasserstein generative adversarial network at the end of the backbone network to enhance the framework performance. VABD can accurately discriminate abnormal video frames from video sequences. Experimental results on UCSD, CUHK Avenue, IITB-Corridor, and ShanghaiTech datasets show that VABD outperforms the state-of-the-art algorithms on abnormal crowd behavior detection. Without data augmentation, our VABD achieves 72.24% in terms of AUC on IITB-Corridor, which surpasses the state-of-the-art methods by nearly 5%.

关键词： Feature extraction Probabilistic logic Video sequences Image reconstruction Anomaly detection Training Optical losses Abnormal crowd behavior detection conditional variational auto-encoder optical flow network motion loss Wasserstein generative adversarial network

来源：评论

学校读者我要写书评

暂无评论

GoRec: A Generative Cold-Start Recommendation Framework 23

GoRec: A Generative Cold-Start Recommendation Framework

引用

31st ACM International Conference on Multimedia (MM)

作者： Bai, Haoyue Hou, Min Wu, Le Yang, Yonghui Zhang, Kun Hong, Richang Wang, Meng Hefei Univ Technol Key Lab Knowledge Engn Big Data Hefei Peoples R China Hefei Univ Technol Key Lab Knowledge Engn Big Data Inst Dataspace Hefei Comprehens Natl Sci Ctr Hefei Peoples R China Hefei Univ Technol Key Lab Knowledge Engn Big Data Inst Artificial Intelligence Hefei Comprehens Natl Sci Ctr Hefei Peoples R China

ISBN: (纸本)9798400701085

Multimedia-based recommendation models learn user and item preference representation by fusing both the user-item collaborative signals and the multimedia content signals. In real scenarios, cold items appear in the test stage without any user interaction record. How to perform cold item recommendation is challenging as the training items and test items have different data distributions. These hybrid preference representations contained auxiliary collaborative signals, so current solutions designed alignment functions to transfer learned hybrid preference representations to cold items. Despite the effectiveness, we argue that they are still limited as these models relied heavily on the manually carefully designed alignment functions, which are easily influenced by the limited item records and noises in the training data. To tackle the above limitations, we propose a Generative cold-start Recommendation (GoRec) framework for multimedia-based new item recommendation. Specifically, we design a conditional variational autoencoder (CVAE) based method that first estimates the underlying distribution of each warm item conditioned on the multimedia content representation. Then, we propose a uniformity-enhanced optimization objective to ensure the latent space of CVAE is more distinguishable and informative. In the inference stage, a generative approach is designed to obtain warm-up new item representations from the latent distribution. Please note that GoRec is applicable to arbitrary recommendation backbones. Extensive experiments on three real datasets and various recommendation backbones verify the superiority of our proposed framework. The code is available at https://***/HaoyueBai98/GoRec.

关键词： recommender system cold start conditional variational auto-encoder

来源：评论

学校读者我要写书评

暂无评论

conditional variational Image Deraining

引用

IEEE TRANSACTIONS ON IMAGE PROCESSING 2020年 29卷 6288-6301页

作者： Du, Yingjun Xu, Jun Zhen, Xiantong Cheng, Ming-Ming Shao, Ling Univ Amsterdam Inst Informat NL-1012 WX Amsterdam Netherlands Nankai Univ Coll Comp Sci TKLNDST Tianjin 300071 Peoples R China Guangdong Univ Petrochem Technol Coll Comp Sci Maoming 525000 Peoples R China Inception Inst Artificial Intelligence Abu Dhabi U Arab Emirates

Image deraining is an important yet challenging image processing task. Though deterministic image deraining methods are developed with encouraging performance, they are infeasible to learn flexible representations for probabilistic inference and diverse predictions. Besides, rain intensity varies both in spatial locations and across color channels, making this task more difficult. In this paper, we propose a conditional variational Image Deraining (CVID) network for better deraining performance, leveraging the exclusive generative ability of conditional variational auto-encoder (CVAE) on providing diverse predictions for the rainy image. To perform spatially adaptive deraining, we propose a spatial density estimation (SDE) module to estimate a rain density map for each image. Since rain density varies across different color channels, we also propose a channel-wise (CW) deraining scheme. Experiments on synthesized and real-world datasets show that the proposed CVID network achieves much better performance than previous deterministic methods on image deraining. Extensive ablation studies validate the effectiveness of the proposed SDE module and CWscheme in our CVID network. The code is available at https://***/Yingjun-Du/VID.

关键词： conditional variational auto-encoder single image deraining spatial attention map channel-wise deraining

来源：评论

学校读者我要写书评

暂无评论

Appliance V-I Trajectory Detection for NILM based on conditional autoencoder

Appliance V-I Trajectory Detection for NILM based on Conditi...

引用

2023 IEEE Sustainable Power and Energy Conference, iSPEC 2023

作者： Yu, Zhiwen Zhao, Ruifeng Jia, Ruiyao Zheng, Wenjie Zhang, Jinjiang Li, Shiming Electric Power Dispatching and Control Center Guangdong Power Grid Guangzhou510600 China The School of Electrical Automation and Information Engineering Tianjin University Tianjin300072 China

ISBN: (纸本)9798350322699

Non-intrusive load monitoring (NILM) analyzes the terminal voltage and total current to offer comprehensive appliance-specific consumption data. Many state-of-the-art NILM models make a crucial assumption that recognized devices within the training set are responsible for triggering switching events. However, continuous addition of new devices diminishes the practical efficacy of existing methods for load monitoring. In this manuscript, we introduce a novel approach based on the conditional variational auto-encoder (CVAE) to address the challenge of classifying familiar appliances and identifying unfamiliar ones by leveraging V-I trajectory characteristics. In our proposed approach, during the training phase, we promote the alignment of capsule features belonging to the same familiar class with a predefined Gaussian distribution, where each class has its own distribution. To achieve this, we adopt the variational autoencoder framework and utilize a collection of Gaussian priors as an estimate for the following distribution. By employing this approach, we can regulate the compactness of features belonging to the same class around the mean of the corresponding Gaussian distributions. This enables us to control the classifier's capability to identify samples from unfamiliar classes. Testing findings on the public dataset illustrate the efficiency of our approach. © 2023 IEEE.

关键词： Capsule system conditional variational auto-encoder Non-intrusive Load Monitoring Unfamiliar Appliances Detection V-I path

来源：评论

学校读者我要写书评

暂无评论

ADS-Cap: A Framework for Accurate and Diverse Stylized Captioning with Unpaired Stylistic Corpora 11th

ADS-Cap: A Framework for Accurate and Diverse Stylized Capti...

引用

11th CCF International Conference on Natural Language Processing and Chinese Computing (NLPCC)

作者： Cheng, Kanzhi Ma, Zheng Zong, Shi Zhang, Jianbing Dai, Xinyu Chen, Jiajun Nanjing Univ Natl Key Lab Novel Software Technol Nanjing Peoples R China

ISBN: (纸本)9783031171208;9783031171192

Generating visually grounded image captions with specific linguistic styles using unpaired stylistic corpora is a challenging task, especially since we expect stylized captions with a wide variety of stylistic patterns. In this paper, we propose a novel framework to generate Accurate and Diverse Stylized Captions (ADS-Cap). Our ADS-Cap first uses a contrastive learning module to align the image and text features, which unifies paired factual and unpaired stylistic corpora during the training process. A conditional variational auto-encoder is then used to automatically memorize diverse stylistic patterns in latent space and enhance diversity through sampling. We also design a simple but effective recheck module to boost style accuracy by filtering style-specific captions. Experimental results on two widely used stylized image captioning datasets show that regarding consistency with the image, style accuracy and diversity, ADS-Cap achieves outstanding performances compared to various baselines. We finally conduct extensive analyses to understand the effectiveness of our method. (Our code is available at https://github. com/njucckevin/ADS-Cap.)

关键词： Stylized image captioning Contrastive learning conditional variational auto-encoder

来源：评论

学校读者我要写书评

暂无评论

Evaluating the process operating state taking into consideration operator interventions with application to a hot rolling mill process

引用

CONTROL ENGINEERING PRACTICE 2024年 155卷

作者： Zhang, Kai Zhang, Xiaowen Peng, Kaixiang Univ Sci & Technol Beijing Sch Automat & Elect Engn Beijing 100083 Peoples R China Univ Sci & Technol Beijing Sch Automat & Elect Engn Key Lab Knowledge Automat Ind Proc Minist Educ Beijing 100083 Peoples R China

In complex industrial processes, process operators often intervene in automatic control systems based on their assessment of the operating state. Traditional operating state evaluation methods do not take into consideration or cannot effectively use this intervention information, and thus may incorrectly evaluate the operating state. In this paper, a convolutional-neural network-conditional variational auto-encoder (CNN-CVAE)-based method for evaluating the operating state is proposed to address this problem. First, the operating-state labels are constructed considering the operator-intervention information. Next, the features of operator-intervention variables (OIVs) are extracted based on CNN, and the obtained probabilities of belonging to different operating states are used as conditional probabilities of CVAE to supervise the feature extraction from the ordinary process data. Finally, both features are fused in a fully connected layer to obtain the predicted operating state. Compared with traditional methods, CNN-CVAE can capture features from both OIVs and process data for evaluating the operating state. The proposed method is validated in areal, hot strip rolling mill process. The results show that the proposed method improves the evaluation accuracy by 54.72% compared with five methods that do not fully use the OIVs.

关键词： Operating state evaluation Operator interventions conditional variational auto-encoder Convolutional neural network Hot strip rolling mill process

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：