检索结果-内蒙古大学图书馆

Interactive Visual Cluster Analysis by Contrastive Dimensionality Reduction

IEEE Transactions on Visualization and computer Graphics 2023年第1期29卷 734-744页

作者： Xia, Jiazhi Huang, Linquan Lin, Weixing Zhao, Xin Wu, Jing Chen, Yang Zhao, Ying Chen, Wei School of Computer Science and Engineering Central South University China Cardiff University United Kingdom State Key Lab of CAD&CG Zhejiang University China Laboratory of Art and Archaeology Image Zhejiang University Ministry of Education China

We propose a contrastive dimensionality reduction approach (CDR) for interactive visual cluster analysis. Although dimensionality reduction of high-dimensional data is widely used in visual cluster analysis in conjunction with scatterplots, there are several limitations on effective visual cluster analysis. First, it is non-trivial for an embedding to present clear visual cluster separation when keeping neighborhood structures. Second, as cluster analysis is a subjective task, user steering is required. However, it is also non-trivial to enable interactions in dimensionality reduction. To tackle these problems, we introduce contrastive learning into dimensionality reduction for high-quality embedding. We then redefine the gradient of the loss function to the negative pairs to enhance the visual cluster separation of embedding results. Based on the contrastive learning scheme, we employ link-based interactions to steer embeddings. After that, we implement a prototype visual interface that integrates the proposed algorithms and a set of visualizations. Quantitative experiments demonstrate that CDR outperforms existing techniques in terms of preserving correct neighborhood structures and improving visual cluster separation. The ablation experiment demonstrates the effectiveness of gradient redefinition. The user study verifies that CDR outperforms t-SNE and UMAP in the task of cluster identification. We also showcase two use cases on real-world datasets to present the effectiveness of link-based interactions. © 2022 IEEE.

关键词： Clustering algorithms

来源：评论

学校读者我要写书评

暂无评论

Effects of View Layout on Situated Analytics for Multiple-View Representations in Immersive Visualization

引用

IEEE Transactions on Visualization and computer Graphics 2023年第1期29卷 440-450页

作者： Wen, Zhen Zeng, Wei Weng, Luoxuan Liu, Yihan Xu, Mingliang Chen, Wei State Key Lab of CAD&CG Zhejiang University China Laboratory of Art and Archaeology Image Zhejiang University Ministry of Education China The Hong Kong University of Science and Technology Guangzhou China Zhengzhou University China

Multiple-view (MV) representations enabling multi-perspective exploration of large and complex data are often employed on 2D displays. The technique also shows great potential in addressing complex analytic tasks in immersive visualization. However, although useful, the design space of MV representations in immersive visualization lacks in deep exploration. In this paper, we propose a new perspective to this line of research, by examining the effects of view layout for MV representations on situated analytics. Specifically, we disentangle situated analytics in perspectives of situatedness regarding spatial relationship between visual representations and physical referents, and analytics regarding cross-view data analysis including filtering, refocusing, and connecting tasks. Through an in-depth analysis of existing layout paradigms, we summarize design trade-offs for achieving high situatedness and effective analytics simultaneously. We then distill a list of design requirements for a desired layout that balances situatedness and analytics, and develop a prototype system with an automatic layout adaptation method to fulfill the requirements. The method mainly includes a cylindrical paradigm for egocentric reference frame, and a force-directed method for proper view-view, view-user, and view-referent proximities and high view visibility. We conducted a formal user study that compares layouts by our method with linked and embedded layouts. Quantitative results show that participants finished filtering- and connecting-centered tasks significantly faster with our layouts, and user feedback confirms high usability of the prototype system. © 2022 IEEE.

关键词： Data visualization

来源：评论

学校读者我要写书评

暂无评论

Character Modelling with Sketches and ODE-Based Shape Creation

引用

Numerical Mathematics(Theory,Methods and Applications) 2023年第3期16卷 720-751页

作者： Ouwen Li Haibin Fu Shaojun Bian Xiaosong Yang Xiaogang Jin Andres Iglesias Algirdas Noreika Lihua You Jian Jun Zhang National Centre for Computer Animation Bournemouth UniversityUnited Kingdom State Key Lab of CAD&CG Zhejiang UniversityChina Department of Applied Mathematics and Computational Sciences University of CantabriaSpain Department of Information Science Toho UniversityJapan Indeform Ltd Lithuania

Character models have enormous applications in *** creation of detailed character models is an important *** paper proposes a new and easy-to-use technique to quickly create detailed character models from *** proposed technique consists of two main components:primitive deformer and shape *** this technique,2D silhouette contours of a character model are drawn or extracted from an image or ***,proper geometric primitives are selected and aligned with the corresponding 2D silhouette *** that,a primitive deformer is used to create a base mesh and three shape generators are used to add 3D details to the base *** primitive deformer and three shape generators are developed from ODE-driven *** primitive deformer deforms the aligned geometric primitives to exactly match the 2D silhouette contours in one view plane and obtains a base mesh of a character model consisting of deformed *** shape generators are used to add 3D details to the base mesh by creating local 3D *** experimental results demonstrate that the new technique can quickly create detailed 3D character models from sketches with few manual *** new technique is physics-based and easy to learn and use.

关键词： Sketch-based character modelling shape generators ordinary differential equations finite difference solution

来源：评论

学校读者我要写书评

暂无评论

NerfCap: Human Performance Capture with Dynamic Neural Radiance Fields

引用

IEEE Transactions on Visualization and computer Graphics 2023年第12期29卷 5097-5110页

作者： Wang, Kangkan Peng, Sida Zhou, Xiaowei Yang, Jian Zhang, Guofeng Nanjing University of Science and Technology Key Lab of Intelligent Percept. and Syst. for High-Dimensional Information of Ministry of Education School of Computer Science and Engineering Jiangsu Nanjing210094 China Zhejiang University State Key Lab of CAD&CG Zijingang Campus Zhejiang Hangzhou310058 China

This paper addresses the challenge of human performance capture from sparse multi-view or monocular videos. Given a template mesh of the performer, previous methods capture the human motion by non-rigidly registering the template mesh to images with 2D silhouettes or dense photometric alignment. However, the detailed surface deformation cannot be recovered from the silhouettes, while the photometric alignment suffers from instability caused by appearance variation in the videos. To solve these problems, we propose NerfCap, a novel performance capture method based on the dynamic neural radiance field (NeRF) representation of the performer. Specifically, a canonical NeRF is initialized from the template geometry and registered to the video frames by optimizing the deformation field and the appearance model of the canonical NeRF. To capture both large body motion and detailed surface deformation, NerfCap combines linear blend skinning with embedded graph deformation. In contrast to the mesh-based methods that suffer from fixed topology and texture, NerfCap is able to flexibly capture complex geometry and appearance variation across the videos, and synthesize more photo-realistic images. In addition, NerfCap can be pre-trained end to end in a self-supervised manner by matching the synthesized videos with the input videos. Experimental results on various datasets show that NerfCap outperforms prior works in terms of both surface reconstruction accuracy and novel-view synthesis quality. © 1995-2012 IEEE.

关键词： Image reconstruction

来源：评论

学校读者我要写书评

暂无评论

Delving into the Reversal Curse: How Far Can Large Language Models Generalize? 38

Delving into the Reversal Curse: How Far Can Large Language ...

引用

38th Conference on Neural Information Processing Systems, NeurIPS 2024

作者： Lin, Zhengkai Fu, Zhihang Liu, Kai Xie, Liang Lin, Binbin Wang, Wenxiao Cai, Deng Wu, Yue Ye, Jieping State Key Lab of CAD&CG Zhejiang University China Alibaba Cloud China College of Biomedical Engineering & Instrument Science Zhejiang University China College of Computer Science and Technology Zhejiang University of Technology China School of Software Technology Zhejiang University China Fullong Inc. China

While large language models (LLMs) showcase unprecedented capabilities, they also exhibit certain inherent limitations when facing seemingly trivial tasks. A prime example is the recently debated "reversal curse", which surfaces when models, having been trained on the fact "A is B", struggle to generalize this knowledge to infer that "B is A". In this paper, we examine the manifestation of the reversal curse across various tasks and delve into both the generalization abilities and the problem-solving mechanisms of LLMs. This investigation leads to a series of significant insights: (1) LLMs are able to generalize to "B is A" when both A and B are presented in the context as in the case of a multiple-choice question. (2) This generalization ability is highly correlated to the structure of the fact "A is B" in the training documents. For example, this generalization only applies to biographies structured in "[Name] is [Description]" but not to "[Description] is [Name]". (3) We propose and verify the hypothesis that LLMs possess an inherent bias in fact recalling during knowledge application, which explains and underscores the importance of the document structure to successful learning. (4) The negative impact of this bias on the downstream performance of LLMs can hardly be mitigated through training alone. These findings offer a novel perspective on interpreting LLMs' generalization through their intrinsic mechanisms and provide insights for developing more effective learning methods. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Kinome-wide polypharmacology profiling of small molecules by multi-task graph isomorphism network approach

引用

Acta Pharmaceutica Sinica B 2023年第1期13卷 54-67页

作者： Lingjie Bao Zhe Wang Zhenxing Wu Hao Luo Jiahui Yu Yu Kang Dongsheng Cao Tingjun Hou Innovation Institute for Artificial Intelligence in Medicine of Zhejiang University College of Pharmaceutical SciencesZhejiang UniversityHangzhou 310058China State Key Lab of CAD&CG Zhejiang UniversityHangzhou 310058China Xiangya School of Pharmaceutical Sciences Central South UniversityChangsha 410013China

Prediction of the interactions between small molecules and their targets play important roles in various applications of drug development,such as lead discovery,drug repurposing and elucidation of potential drug side ***,a variety of machine learning-based models have been developed to predict these *** this study,a model called auxiliary multi-task graph isomorphism network with uncertainty weighting(AMGU)was developed to predict the inhibitory activities of small molecules against 204 different kinases based on the multi-task Graph Isomorphism Network(MT-GIN)with the auxiliary learning and uncertainty weighting *** calculation results illustrate that the AMGU model outperformed the descriptor-based models and state-of-the-art graph neural networks(GNN)models on the internal test ***,it also exhibited much better performance on two external test sets,suggesting that the AMGU model has enhanced generalizability due to its great transfer learning ***,a naÏve model-agnostic interpretable method for GNN called edges masking was devised to explain the underlying predictive mechanisms,and the consistency of the interpretability results for 5typical epidermal growth factor receptor(EGFR)inhibitors with their structure-activity relationships could be ***,a free online web server called KIP was developed to predict the kinomewide polypharmacology effects of small molecules(http://***/kip).

关键词： Kinome-wide polypharmacology Machine learning Kinases Graph neural networks Artificial intelligence

来源：评论

学校读者我要写书评

暂无评论

Cloud-based Fusion Rendering for AR Applications 3

Cloud-based Fusion Rendering for AR Applications

引用

3rd IEEE International Conference on Digital Twins and Parallel Intelligence, DTPI 2023

作者： Tang, Jing Zhou, Yuxi Chen, Zhenyu Chen, Cheng Xiong, Gang Zhang, Hongxin Zhejiang University State Key Lab of Cad&cg Hangzhou China Zhejiang University College of Computer Science and Technology Hangzhou China Institute of Automation Chinese Academy of Sciences Beijing China

ISBN: (纸本)9798350318470

To create digital twins, it is necessary to represent real objects or scenes in a virtual form. With the rapid development of Augmented Reality (AR) technology, its applications in various fields are becoming increasingly extensive. However, in traditional AR applications, the client has high performance requirements for device compute-intensive tasks, which cannot be met on low-performance devices. To solve this issue, in this paper, we propose and implement a novel digital twin method for AR technology, which is based on a cloud fusion platform. The approach includes deploying the Neural Radiance Field (NeRF) on a cloud server for high-quality model rendering and integrating simultaneous localization and mapping (SLAM) with AR visualization on the client side. This design effectively separates compute-intensive tasks from client-side rendering tasks, reducing the performance requirements of client devices and enhancing overall scalability and availability of the system. Practical application and tests show that the proposed AR system based on the cloud fusion platform performs well in terms of performance, stability, and scalability. This paper provides detailed discussions on the relevant technical methods and approaches, and showcases examples that serve as a foundation for the implementation of digital twins across diverse fields. © 2023 IEEE.

关键词： Augmented reality

来源：评论

学校读者我要写书评

暂无评论

HetVis: A Visual Analysis Approach for Identifying Data Heterogeneity in Horizontal Federated Learning

引用

IEEE Transactions on Visualization and computer Graphics 2023年第1期29卷 310-319页

作者： Wang, Xumeng Chen, Wei Xia, Jiazhi Wen, Zhen Zhu, Rongchen Schreck, Tobias TMCC CS Nankai University China State Key Lab of CAD&CG Zhejiang University China Laboratory of Art and Archaeology Image Zhejiang University Ministry of Education China School of Computer Science and Engineering Central South University China Graz University of Technology Austria

Horizontal federated learning (HFL) enables distributed clients to train a shared model and keep their data privacy. In training high-quality HFL models, the data heterogeneity among clients is one of the major concerns. However, due to the security issue and the complexity of deep learning models, it is challenging to investigate data heterogeneity across different clients. To address this issue, based on a requirement analysis we developed a visual analytics tool, HetVis, for participating clients to explore data heterogeneity. We identify data heterogeneity through comparing prediction behaviors of the global federated model and the stand-alone model trained with local data. Then, a context-aware clustering of the inconsistent records is done, to provide a summary of data heterogeneity. Combining with the proposed comparison techniques, we develop a novel set of visualizations to identify heterogeneity issues in HFL. We designed three case studies to introduce how HetVis can assist client analysts in understanding different types of heterogeneity issues. Expert reviews and a comparative study demonstrate the effectiveness of HetVis. © 2022 IEEE.

关键词： Cluster analysis

来源：评论

学校读者我要写书评

暂无评论

Research on Emotional Classification and Literary Narrative Visualization Based on Graph Convolutional Neural Network 12th

Research on Emotional Classification and Literary Narrative ...

引用

12th International Conference on Image and Graphics, ICIG 2023

作者： Zhuo, Shi Meng, Wang Wei, Chen Xiaonan, Luo Guilin University of Electronic Science and Technology Guilin China State Key Lab of CAD&CG Zhejiang University Hangzhou China

ISBN: (纸本)9783031463105

Watching a movie for three minutes has become a popular term in contemporary life, and people also hope to spend less time understanding the emotional direction and general plot of long novels. Currently, research on emotional analysis in novels mainly focuses on foreign languages. This article introduces Graph Convolutional Neural networks into the text emotion classification of novels. Text GCN is used to conduct emotion analysis on Ba Jin’s novel "The Rapids Trilogy", and iStoryline is used to visualize the storyline of literary works. Visual coding adjustments are made based on the results of emotion analysis in the generated storyline, where line aggregation and dispersion represent whether the characters appear in the same scene, The brighter the color of the lines, the more positive the emotions, making the visualization more vivid, full, and beautiful. The experimental results indicate that this method enables readers to easily grasp the overall direction of the novel’s plot and have a clearer understanding of the emotional changes of the characters. In addition, the accuracy of the Text GCN model is superior to traditional sentiment classification methods on Chinese datasets. This method provides a new approach for emotional analysis and narrative visualization research in literary works, and can be extended to other fields of digital entertainment. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023.

关键词： Visualization

来源：评论

学校读者我要写书评

暂无评论

Incremental Object Detection With Image-Level labels

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2024年第5期5卷 2331-2341页

作者： Yu, Qianzi Zhu, Kai Wang, Wen Cao, Yang Kang, Yu University of Science and Technology of China Department of Automation School of Information Science and Technology Hefei230026 China Hefei Comprehensive National Science Center Institute of Artificial Intelligence Hefei230039 China Zhejiang University College of Computer Science and Technology State Key Lab of CAD&CG Hangzhou310027 China University of Science and Technology of China School of Information Science and Technology Hefei230026 China

Incremental object detection (IOD) aims to achieve simultaneous prediction of old and new samples on localization and classification when new concepts are provided. It is a challenging task due to the need for a joint interpretation of semantic and spatial information. While the existing work accomplishes the detector generalization with the help of annotated sample classes and bounding boxes, this article presents one astonishing finding on the unnecessity of new bounding boxes, which will significantly reduce the annotation cost and condition constraints in IOD. To enable the incremental detection process with image-level labels, we propose a multibranch decoupling scheme in which the representation, classification, and regression branches are customized to accommodate new concepts with class labels with different semantic levels. The regression branch is first frozen to ensure generalizable localization while maintaining the stability of representation optimization. Then, the representation branch is gradient-inverted to estimate the distribution drift from the old object-level set to the new image-level one, improving the effectiveness of feature updates in the absence of homogeneous semantic and spatial supervision. Finally, the classification branch is calibrated to balance the contribution of different images to the gradient, in which the frequency ratios of old instances and new images are calculated as a postprocessing factor for the logits margin. Extensive experiments on two standard incremental detection benchmarks demonstrate the above-par recognition and localization performance while reducing the burden on new class annotation, outperforming state-of-the-art methods by 2% and 3%, respectively. © 2023 IEEE.

关键词： Feature extraction

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：