检索结果-内蒙古大学图书馆

FireSeg: A weakly supervised fire segmentation framework via pre-trained latent diffusion models

EXPERT SYSTEMS WITH APPLICATIONS 2025年 275卷

作者： Zhang, Wei Zheng, Hongtao Li, Weiran pan, shuwen Liu, Yan Zhejiang Tech Inst Econ Digital Informat Technol Sch Hangzhou 310018 Peoples R China Univ Colorado Dept Comp Sci Boulder CO USA Hangzhou City Univ Sch Informat & Elect Engn Hangzhou 310015 Peoples R China Suzhou Caiju Intelligent Technol Co Ltd Suzhou 215332 Peoples R China Zhejiang Univ Sch Control Sci & Engn Hangzhou 310058 Peoples R China

Accurate fire segmentation is crucial for emergency fire response. However, current deep learning-based fire segmentation methods heavily rely on large-scale annotated data, and the scarcity of high-quality public datasets limits their application in practical, diverse scenarios. To address this, we propose an innovative weakly-supervised fire segmentation framework, FireSeg, which consists of two core modules: 1) a feature extraction module based on a pre-trained latent diffusion model, generating cross-attention maps through diffusion inversion and textual prompts;and 2) a Transformer decoder-based Flame-Decoder module, which progressively decodes the attention maps and outputs segmentation masks by incorporating Dynamic Focus- aware Positional Queries (DFPQ) and a High-Resolution Cross-Attention (HRCA) mechanism. Experimental results demonstrate that FireSeg significantly outperforms other models on three fully supervised datasets covering different fire scenarios. For example, on the BowFire dataset, FireSeg outperforms the secondbest model by 4.10%;on the Corsican dataset, it exceeds the second-best model by 3.83%;and on the FLAME dataset, it surpasses the second-best model by 4.24%. Additionally, FireSeg also shows a significant improvement over the second-best model on the weakly supervised FireDM dataset, which contains diverse scenes but lower-quality annotations. Notably, when only 20% of the weakly supervised data is used, FireSeg achieves a 9.23% increase in IoU. These results indicate that FireSeg not only provides excellent segmentation performance in fully supervised settings but also demonstrates strong adaptability and effectiveness in weakly supervised scenarios with limited labeled data.

关键词： Fire Segmentation Latent diffusion model Textual information Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Hierarchical supervisions with two-stream network for Deepfake detection

引用

PATTERN RECOGNITION LETTERS 2023年第1期172卷 121-127页

作者： Liang, Yufei Wang, Mengmeng Jin, Yining pan, shuwen Liu, Yong Zhejiang Univ Coll Control Sci & Engn Lab Adv Percept Robot & Intelligent Learning Hangzhou Peoples R China Univ Alberta Dept Elect & Comp Engn Edmonton AB Canada Zhejiang Univ City Coll Sch Informat & Elect Engn Hangzhou Peoples R China

Recently, the quality of face generation and manipulation has reached impressive levels, making it diffi-cult even for humans to distinguish real and fake faces. At the same time, methods to distinguish fake faces from reals came out, such as Deepfake detection. However, the task of Deepfake detection remains challenging, especially the low-quality fake images circulating on the Internet and the diversity of face generation methods. In this work, we propose a new Deepfake detection network that could effectively distinguish both high-quality and low-quality faces generated by various generation methods. First, we design a two-stream framework that incorporates a regular spatial stream and a frequency stream to handle the low-quality problem since we find that the frequency domain artifacts of low-quality images will be preserved. Second, we introduce hierarchical supervisions in a coarse-to-fine manner, which con-sists of a coarse binary classification branch to classify reals and fakes and a five-category classification branch to classify reals and four different types of fakes. Extensive experiments have proved the effec-tiveness of our framework on several widely used datasets.& COPY;2023 Published by Elsevier B.V.

关键词： Deepfake detection Frequency domain Two stream Coarse to fine

来源：评论

学校读者我要写书评

暂无评论

Omni-Frequency Channel-Selection Representations for Unsupervised Anomaly Detection

引用

IEEE TRANSACTIONS ON IMAGE PROCESSING 2023年 32卷 4327-4340页

作者： Liang, Yufei Zhang, Jiangning Zhao, Shiwei Wu, Runze Liu, Yong pan, shuwen Zhejiang Univ Coll Control Sci & Engn Lab Adv Percept Robot & Intelligent Learning Hangzhou 310027 Peoples R China NetEase Fuxi AI Lab Hangzhou 310012 Peoples R China Hangzhou City Univ Sch Informat & Elect Engn Discipline Control Sci & Engn Hangzhou 310015 Peoples R China

Density-based and classification-based methods have ruled unsupervised anomaly detection in recent years, while reconstruction-based methods are rarely mentioned for the poor reconstruction ability and low performance. However, the latter requires no costly extra training samples for the unsupervised training that is more practical, so this paper focuses on improving reconstruction-based method and proposes a novel Omni-frequency Channel-selection Reconstruction (OCR-GAN) network to handle sensory anomaly detection task in a perspective of frequency. Concretely, we propose a Frequency Decoupling (FD) module to decouple the input image into different frequency components and model the reconstruction process as a combination of parallel omni-frequency image restorations, as we observe a significant difference in the frequency distribution of normal and abnormal images. Given the correlation among multiple frequencies, we further propose a Channel Selection (CS) module that performs frequency interaction among different encoders by adaptively selecting different channels. Abundant experiments demonstrate the effectiveness and superiority of our approach over different kinds of methods, e.g., achieving a new state-of-theart 98.3 detection AUC on the MVTec AD dataset without extra training data that markedly surpasses the reconstruction-based baseline by +38.11. and the current SOTA method by +0.31.. The source code is available in the additional materials.

关键词： Index Terms- Anomaly detection omni-frequency decoupling unsupervised learning reconstruction-based network

来源：评论

学校读者我要写书评

暂无评论

Development and Algorithm Design of Transportation Countermeasure Robot System 5

Development and Algorithm Design of Transportation Counterme...

引用

5th International Symposium on Computer Engineering and Intelligent Communications, ISCEIC 2024

作者： Gao, Tianyang Liu, Yan Shao, Linye Ma, Zhuoyuan pan, shuwen Li, Yanjun School of Information And Electrical Engineering Hangzhou City University Hang Zhou China

ISBN: (纸本)9798331518677

In response to the design requirements and competition rules of the '8th Zhejiang College Student Robot Competition', we independently developed a set of robot systems. The system takes the STM32F103RCT6 microcontroller as the core, controls the motor and the robotic arm, and realizes the functions of moving, grabbing and placing. Equipped with seven-way and single-channel grayscale sensors to ensure straight line inspection, intersection detection and accurate positioning. The software part covers the robot's motion control, robotic arm control, robot flexible start and stop, path planning and design algorithms, the PID algorithm is used to achieve the precise control of the robot, and the A ∗ algorithm is used for the simplest path planning. Based on the research of game theory, five sets of schemes were designed, and the matrix keyboard was designed to realize the choice of different strategies in the game. The test experiment results show that the ranking competition is about 50 seconds, and the stability of the robot is found to be higher than 95%, the robot system has been applied to the eighth college student robot competition in Zhejiang Province, through a round of ranking competition, three rounds of group stage and two rounds of knockout competition, and finally won the second prize of Zhejiang Province, which also shows that the five sets of strategies are effective and feasible. The transportation robot has a series of advantages such as low cost and strong scalability, which can provide reference for various robot competitions and related practical projects for college students. At the same time, we also put forward effective solutions to a series of problems encountered in the process of practice. © 2024 IEEE.

关键词： Robotic arms

来源：评论

学校读者我要写书评

暂无评论

Fine-Grained Image Editing Using ControlNet: Expanding Possibilities in Visual Manipulation 20th

Fine-Grained Image Editing Using ControlNet: Expanding Possi...

引用

20th International Conference on Intelligent Computing (ICIC)

作者： Xu, Longfei Huang, Hongbo Zhao, Yushuang pan, shuwen Zheng, Yaolin Yan, Xiaoxu Huang, Linkai Wu, Lishan Beijing Informat Sci & Technol Univ Beijing Peoples R China

ISBN: (纸本)9789819755967;9789819755974

In recent years, diffusion probabilistic models have emerged as a hot topic in computer vision. Image creation programs such as Imagen, Latent Diffusion Models, and Stable Diffusion have shown outstanding generative powers, sparking considerable community discussions. They frequently, however, lack the ability to precisely modify real-world images. In this paper, we propose a novel ControlNet-based image editing framework that enables alteration of real images based on pose maps, scribbling maps, and other features without the need for training or fine-tuning. Given a guiding image as input, we edit the initial noise generated from the guiding image to influence the generation process. Then features extracted from the guiding image are directly injected into the generation process of the translated image. We also construct a classifier guidance based on strong correspondences between intermediate features of the ControlNet branches. The editing signals are converted into gradients to guide the sampling direction. At the end of this paper, we demonstrate high-quality results of our proposed model in image editing tasks.

关键词： Diffusion Probabilistic Model Controlnet Image Editing

来源：评论

学校读者我要写书评

暂无评论

Analyses of Shanghai Disneyland IP localization Strategy And Its Risks from the Perspective of Lina Bell-Based on SWOT Analysis Method

Analyses of Shanghai Disneyland IP localization Strategy And...

引用

作者： shuwen pan School of Foreign Language

'Lina Bell' is a new IP image that was exposed by Shanghai Disney in September *** 'new role' to 'online celebrity with a huge fan base' in just one month,its popularity overwhelmed other famou...

'Lina Bell' is a new IP image that was exposed by Shanghai Disney in September *** 'new role' to 'online celebrity with a huge fan base' in just one month,its popularity overwhelmed other famous Disney IPs,and the price of its doll has ***,in early December 2021,'Lina Bell' was stuck in a 'image crisis',and there are still risks behind Disney's IP *** paper analyses Disney's IP localization strategy through a SWOT method and proposes possible countermeasures to address the potential *** study finds that the history of Disney's Duffy family is the epitome of Disney's IP localization strategy,and that Lina Bell is the best product of the IP localization strategy;Disney's strengths are its rich experience in IP operation and complete industry chain,while its weaknesses are mainly the lack of a true understanding of the target market's culture;the external opportunities for Disney are the huge potential of the target market and the fact that Disney IP is in line with the pop culture trends of the target market,while the external threats are mainly public opinion and competition from similar *** paper suggests that Disney should strengthen its research into the culture of its target markets and truly understand the local cultural habits;strengthen the training of its staff;standardize the behaviour of each staff member in line with the brand's mission and increase the local core competencies of the Disney IP.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Machine learning-derived diagnostic model of epithelial ovarian cancer based on gut microbiome signatures

引用

JOURNAL OF TRANSLATIONAL MEDICINE 2025年第1期23卷 1-23页

作者： Chen, Cheng Deng, Chengyuan Li, Yanwen He, Shuguang Liu, Yunhong pan, shuwen Xu, Wenqian Fang, Lu Zhu, Yixi Wang, Yingying Jiang, Xiaoxin Hunan Prov Directly Affiliated TCM Hosp Affiliated Hosp 1 Hunan Tradit Chinese Med Coll Dept Clin Lab Zhuzhou Peoples R China Peoples Hosp Longhua Dept Clin Lab Shenzhen Peoples R China Hunan Tradit Chinese Med Coll Zhuzhou Peoples R China Child Healthcare Hosp Dept Clin Lab Shenzhen Longhua Matern Shenzhen Peoples R China Changsha Ctr Dis Prevent & Control Changsha Peoples R China Xuzhou First Peoples Hosp Xuzhou Peoples R China

BackgroundPrior studies have elucidated that alterations in gut microbiota are associated with a spectrum of tumors and metabolic disorders. However, the diagnostic value of gut microbiota in epithelial ovarian cancer remains insufficiently *** total of 34 patients with a diagnosis of epithelial ovarian cancer (EOC), 15 patients with benign ovarian tumors (TB), and 30 healthy volunteers (NOR) were enrolled in this study. Fecal samples were collected, followed by sequencing of the V3-V4 region of the 16S rRNA gene. The clinical data and pathological characteristics were comprehensively recorded for further analysis, PICRUSt2 was utilized to conduct an analysis of microbial functional predictions, WGCNA networks were constructed by integrating microbiome and clinical data. LEfSe analysis was employed to identify microbial diagnostic markers, LASSO and SVM analyses were used to screen microbial diagnostic markers in conjunction with the Cally index, to establish a Microbial-Cally diagnostic model. Bootstrap resampling was utilized for the internal validation of the model, whereas the Hosmer-Lemeshow test and decision curve analysis (DCA) were employed to evaluate the diagnostic performance of the model. Plasma samples were subjected to untargeted metabolomics profiling, followed by differential analysis to identify key metabolites that are significantly altered in epithelial ovarian cancer. At the same time, Spearman correlation analysis was used to study the association between key microbiota and differential metabolites. The supernatants from Escherichia coli and Bifidobacterium cultures were co-cultured with SKOV3 cells. Cell proliferation, migration, and invasion were evaluated using Cell Counting Kit-8 (CCK-8) assay, Transwell migration and invasion assays. Apoptosis was assessed by flow cytometry analysis of fluorescence signals from Annexin V and propidium iodide (PI) *** to Nor and TB populations, individuals diagnosed wit

关键词： Epithelial ovarian cancer Gut microbiota Enterotype diagnostic model

来源：评论

学校读者我要写书评

暂无评论

Enhanced Thermal Modeling of Electric Vehicle Motors Using a Multihead Attention Mechanism

引用

ENERGIES 2024年第12期17卷 2976页

作者： Ji, Feifan Huang, Chenglong Wang, Tong Li, Yanjun pan, shuwen Hangzhou City Univ Sch Informat & Elect Engn Hangzhou 310015 Peoples R China ZheJiang Dahua Technol CO Ltd Hangzhou 310000 Peoples R China

The rapid advancement of electric vehicles (EVs) accentuates the criticality of efficient thermal management systems for electric motors, which are pivotal for performance, reliability, and longevity. Traditional thermal modeling techniques often struggle with the dynamic and complex nature of EV operations, leading to inaccuracies in temperature prediction and management. This study introduces a novel thermal modeling approach that utilizes a multihead attention mechanism, aiming to significantly enhance the prediction accuracy of motor temperature under varying operational conditions. Through meticulous feature engineering and the deployment of advanced data handling techniques, we developed a model that adeptly navigates the intricacies of temperature fluctuations, thereby contributing to the optimization of EV performance and reliability. Our evaluation using a comprehensive dataset encompassing temperature data from 100 electric vehicles illustrates our model's superior predictive performance, notably improving temperature prediction accuracy.

关键词： thermal model electric vehicle motors multihead attention mechanism temperature prediction

来源：评论

学校读者我要写书评

暂无评论

Efficient Backup Link Scheme of ECHO Protocol for Mobile Ad Hoc Networks

Efficient Backup Link Scheme of ECHO Protocol for Mobile Ad ...

引用

2022 IEEE International Conference on Signal Processing, Communications and Computing, ICSPCC 2022

作者： Deng, Huoyun Ou, Jinghang Yao, Chenlong Lu, Xiangqun Tan, Weijie pan, shuwen Wang, Jing Beijing University of Posts and Telecommunications School of Information and Communication Engineering Beijing China Beijing University of Posts and Telecommunications School of Computer Science Beijing China Guizhou University Guizhou Big Data Academy Guiyang China Beijing Information Science and Technology University Computer School Beijing China Guizhou University College of Big Data and Information Engineering Guiyang China

ISBN: (数字)9781665469722

ISBN: (纸本)9781665469722

Ad hoc wireless mesh network is widely used in emergency rescue, military battlefield and smart city due to its flexible networking and easy to deployment. However, traditional routing protocols have the problem of high packet loss rate in the case of poor link quality. To address this issue, this paper proposes an optimization router scheme considering the "backup path"- ECHOBL protocol. In the proposed protocol, we set the critical nodes according to the signal-to-noise ratio(SNR) conditions of the link to improve the communication reliability. Moreover, we introduce the networking mode and optimization point. Finally, we prove correctness and integrity of the proposed protocol. Field experimental results show that, compared with ECHO protocol and Flooding protocol, ECHOBL protocol has lower packet loss rate and better performance under various packet frequencies. © 2022 IEEE.

关键词： Mobile ad hoc networks

来源：评论

学校读者我要写书评

暂无评论

Development and Algorithm Design of Transportation Countermeasure Robot System

Development and Algorithm Design of Transportation Counterme...

引用

Computer Engineering and Intelligent Communications (ISCEIC), International Symposium on

作者： Tianyang Gao Yan Liu Linye Shao Zhuoyuan Ma shuwen pan Yanjun Li School of Information And Electrical Engineering Hangzhou City University Hang Zhou China

ISBN: (数字)9798331518677

ISBN: (纸本)9798331518684

In response to the design requirements and competition rules of the “8th Zhejiang College Student Robot Competition”, we independently developed a set of robot systems. The system takes the STM32F103RCT6 microcontroller as the core, controls the motor and the robotic arm, and realizes the functions of moving, grabbing and placing. Equipped with seven-way and single-channel grayscale sensors to ensure straight line inspection, intersection detection and accurate positioning. The software part covers the robot's motion control, robotic arm control, robot flexible start and stop, path planning and design algorithms, the PID algorithm is used to achieve the precise control of the robot, and the A * algorithm is used for the simplest path planning. Based on the research of game theory, five sets of schemes were designed, and the matrix keyboard was designed to realize the choice of different strategies in the game. The test experiment results show that the ranking competition is about 50 seconds, and the stability of the robot is found to be higher than 95%, the robot system has been applied to the eighth college student robot competition in Zhejiang Province, through a round of ranking competition, three rounds of group stage and two rounds of knockout competition, and finally won the second prize of Zhejiang Province, which also shows that the five sets of strategies are effective and feasible. The transportation robot has a series of advantages such as low cost and strong scalability, which can provide reference for various robot competitions and related practical projects for college students. At the same time, we also put forward effective solutions to a series of problems encountered in the process of practice.

关键词： Software algorithms Transportation Robot sensing systems Software Path planning End effectors Software reliability Sensors Servomotors Robots

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：