检索结果-内蒙古大学图书馆

International Conference on computer Vision (ICCV)

作者： Yin-Yin He Jianxin Wu Xiu-Shen Wei State Key Laboratory for Novel Software Technology Nanjing University China School of Computer Science and Engineering Nanjing University of Science and Technology China

ISBN: (纸本)9781665428132

We tackle the long-tailed visual recognition problem from the knowledge distillation perspective by proposing a Distill the Virtual Examples (DiVE) method. Specifically, by treating the predictions of a teacher model as virtual examples, we prove that distilling from these virtual examples is equivalent to label distribution learning under certain constraints. We show that when the virtual example distribution becomes flatter than the original input distribution, the under-represented tail classes will receive significant improvements, which is crucial in long-tailed recognition. The proposed DiVE method can explicitly tune the virtual example distribution to become flat. Extensive experiments on three benchmark datasets, including the large-scale iNaturalist ones, justify that the proposed DiVE method can significantly outperform state-of-the-art methods. Further-more, additional analyses and experiments verify the virtual example interpretation, and demonstrate the effectiveness of tailored designs in DiVE for long-tailed problems.

关键词： Visualization computer vision Predictive models Benchmark testing

来源：评论

学校读者我要写书评

暂无评论

Using Mobile Sinks in WSN: Computational Complexity and a Theoretical Bound

引用

Chinese Journal of Electronics 2023年第1期20卷 147-150页

作者： Yu Gu Yusheng Ji Hongyang Chen Baohua Zhao Department of Computer Science University of Science and Technology of China Hefei China State Key Laboratory of Networking and Switching Technology Beijing China Information Systems Architecture Science Research Division NII Tokyo Japan Institute of Industrial Science University of Tokyo Tokyo Japan

In this paper, we study the lifetime optimization problem in wireless sensor networks using mobile sink nodes. This problem is inherently difficult since we need to consider both sink scheduling and data routing. Through a simple case study we develop a novel notation named the Placement pattern (PP) to bound traffic patterns with candidate locations. This significantly decreases the number of elements needed to be scheduled. Based on the PP, we mathematically formulate this optimization problem as a Mixed-integer non-linear programming (MINLP), which is very tough and time consuming to solve. By proving that the problem is NP-complete, we point out that instead of seeking an optimal algorithm, heuristic algorithms, especially those with performance guarantee, would be much more desirable to develop. Furthermore, in order to help identify performance gains of heuristic algorithms proposed in the future, we develop a Linear programming (LP) formulation which serves as an upper bound by adopting a reformulation and relaxation technique.

关键词： Wireless sensor networks Upper bound Heuristic algorithms Traffic control Programming Performance gain Routing

来源：评论

学校读者我要写书评

暂无评论

Affine particle-in-cell method for two-phase liquid simulation

引用

Virtual Reality & Intelligent Hardware 2021年第2期3卷 105-117页

作者： Luan LYU Wei CAO Enhua WU Zhixin YANG Faculty of Science and Technology University of MacaoMacao 999078China State Key Laboratory of Internet of Things for Smart City University of MacaoMacao 999078China State Key Laboratory of Computer Science Institute of SoftwareChinese Academy of Sciences(CAS)&University of CASBeijing 100190China Guangzhou Greater Bay Area Virtual Reality Research Institute Guangzhou 510700China

Background The interaction of gas and liquid can produce many interesting phenomena,such as bubbles rising from the bottom of the *** simulation of two-phase fluids is a challenging topic in computer *** animate the interaction of a gas and liquid,MultiFLIP samples the two types of particles,and a Euler grid is used to track the interface of the liquid and ***,MultiFLIP uses the fluid implicit particle(FLIP)method to interpolate the velocities of particles into the Euler grid,which suffer from additional noise and *** To solve the problem caused by fluid implicit particles(FLIP),we present a novel velocity transport technique for two individual particles based on the affine particle-in-cell(APIC)***,we design a weighed coupling method for interpolating the velocities of liquid and gas particles to the Euler grid such that we can apply the APIC method to the simulation of a two-phase ***,we introduce a narrowband method to our system because MultiFLIP is a time-consuming approach owing to the large number of *** Experiments show that our method is well integrated with the APIC method and provides a visually credible two-phase fluid *** The proposed method can successfully handle the simulation of a two phase fluid.

关键词： Fluid simulation Two-Phase flow Affine particle-in-cell method

来源：评论

学校读者我要写书评

暂无评论

Segment Any Organoid: A Generalist Algorithm for High-Throughput Drug Screening of Organoid

SSRN

引用

SSRN 2024年

作者： Gu, Zhongze Du, Xuan Li, Yuchen Song, Jiaping Zhang, Zilin Zhang, Jing Leong, Kam Li, Yanhui Chen, Zaozao State Key Laboratory of Digital Medical Engineering School of Biological Science and Medical Engineering Southeast University Jiangsu Nanjing211189 China State Key Laboratory for Novel Software Technology Nanjing University Jiangsu Nanjing210008 China Jiangsu Avatarget Biotechnology Co. Ltd. Suzhou China Anhui Province Key Laboratory of Functional Agriculture and Functional Food Anhui Science and Technology University Chuzhou239200 China Department of Biomedical Engineering Columbia University New YorkNY10032 United States

Tumor organoids have higher sensitivity, heterogeneity, and stability than traditional drug screening models, enabling a more accurate recapitulation of tumor characteristics. Therefore, organoids play a crucial role in drug screening. Usually, measuring the activity of organoids is one of the main methods to evaluate drug efficacy. However, traditional biochemical methods face significant measurement errors, high costs, slow speed, invasiveness, and limited intuitiveness, severely hindering their application in large-scale preclinical drug screening. In contrast, morphological changes in organoids observed under bright-field microscopy offer a more intuitive and non-invasive means to assess drug efficacy. This paper introduces an algorithm to segment any organoid (SAO) under bright-field microscopy, which has the characteristics of accuracy, high throughput, and strong intuitiveness. It can overcome various imaging artefacts during organoid culture and imaging. In addition, we further verified that the organoid growth rate calculated by the algorithm is highly consistent with the results of organoid activity measurement in the drug sensitivity experiment. We propose that SAO has the potential to serve as a universal analysis tool for organoids, especially in the automated or semi-automated evaluation of large-scale drug screening. The promotion and application of this tool are expected to provide new technical support for the research of precision medicine and personalized treatment, while also promoting the broader utilization of organoid models in drug development and disease research. © 2024, The Authors. All rights reserved.

关键词： Image segmentation

来源：评论

学校读者我要写书评

暂无评论

Student-Teacher Feature Pyramid Matching for Anomaly Detection 32

Student-Teacher Feature Pyramid Matching for Anomaly Detecti...

引用

32nd British Machine Vision Conference, BMVC 2021

作者： Wang, Guodong Han, Shumin Ding, Errui Huang, Di State Key Laboratory of Software Development Environment Beihang University Beijing China School of Computer Science and Engineering Beihang University Beijing China Department of Computer Vision Technology Baidu Inc. Beijing China

Anomaly detection is a challenging task and usually formulated as an one-class learning problem for the unexpectedness of anomalies. This paper proposes a simple yet powerful approach to this issue, which is implemented in the student-teacher framework for its advantages but substantially extends it in terms of both accuracy and efficiency. Given a strong model pre-trained on image classification as the teacher, we distill the knowledge into a single student network with the identical architecture to learn the distribution of anomaly-free images and this one-step transfer preserves the crucial clues as much as possible. Moreover, we integrate the multi-scale feature matching strategy into the framework, and this hierarchical feature matching enables the student network to receive a mixture of multi-level knowledge from the feature pyramid under better supervision, thus allowing to detect anomalies of various sizes. The difference between feature pyramids generated by the two networks serves as a scoring function indicating the probability of anomaly occurring. Due to such operations, our approach achieves accurate and fast pixel-level anomaly detection. Very competitive results are delivered on the MVTec anomaly detection dataset, superior to the state of the art ones. © 2021. The copyright of this document resides with its authors.

关键词： Anomaly detection

来源：评论

学校读者我要写书评

暂无评论

An empirical study on issue knowledge transfer from python to R for machine learning software 32

An empirical study on issue knowledge transfer from python t...

引用

32nd International Conference on software Engineering and Knowledge Engineering, SEKE 2020

作者： Huang, Wenchin Ji, Zhenlan Li, Yanhui State Key Laboratory for Novel Software Technology Nanjing University China Department of Computer Science and Technology Nanjing University China School of Management and Engineering Nanjing University China

ISBN: (纸本)1891706500

Background: With the blowout of programming languages, developers employ different languages to solve similar problems (e.g., to implement machine learning algorithms) separately and frequently, which gives rise to knowledge transfer across different language development. Since GitHub provides an issue tracking system for developers and users to follow with issues, knowledge about how to deal with issues is a main part of available knowledge on GitHub. Such issue knowledge could be directly transferred to help developers handle new issues on current projects from similar projects in different languages. Aims: Inspired by a large amount of developed and developing machine learning software written in Python and R on GitHub, we aim to discover how much issue knowledge can be transferred from Python projects to R projects. Method: We investigate totally 1161 issues from 15 popular machine learning projects in R and 7496 issues from Scikit-Learn in Python on GitHub. After computing the text similarity between issues from R and Python projects, we match top 5 similar Scikit-Learn issues for each R issue and manually judge 1161×5 issue-pairs to label and group them. Results: We observe that a) 13% (149/1161) of R issues can refer to related Python issues;b) 47% (71/149) of related R issues can be linked to Python issues by the text mining technique BM25 at the very early stage;c) 83% (124/149) of related Python issues support code and description about the similar machines learning problems;d) reference knowledge is considered as the most useful knowledge from Python issues. Conclusion: We put forward the following suggestions: a) referring to the corresponding cross languages issues is an efficient way for developers, especially there is the lack of related information in current language;b) the text mining technique BM25 is helpful for developers to start earlier for searching similar issues cross languages. © 2020 Knowledge Systems Institute Graduate School. All rights

关键词： Python

来源：评论

学校读者我要写书评

暂无评论

MiM: Mask in Mask Self-Supervised Pre-Training for 3D Medical Image Analysis

引用

IEEE Transactions on Medical Imaging 2025年

作者： Zhuang, Jiaxin Wu, Linshan Wang, Qiong Fei, Peng Vardhanabhuti, Varut Luo, Lin Chen, Hao Hong Kong University of Science and Technology Departments of Computer Science and Engineering Hong Kong Chinese Academy of Sciences Shenzhen Institutes of Advanced Technology China Huazhong University of Science and Technology School of Optical and Electronic Information Wuhan National Laboratory for Optoelectronics Wuhan430074 China The University of Hong Kong Department of Diagnostic Radiology Hong Kong Peking University College of Engineering Beijing China Hong Kong University of Science and Technology Department of Computer Science and Engineering Department of Chemical and Biological Engineering State Key Laboratory of Molecular Neuroscience Hong Kong HKUST Shenzhen-Hong Kong Collaborative Innovation Research Institute Futian Shenzhen China

The Vision Transformer (ViT) has demonstrated remarkable performance in Self-Supervised Learning (SSL) for 3D medical image analysis. Masked AutoEncoder (MAE) for feature pre-training can further unleash the potential of ViT on various medical vision tasks. However, due to large spatial sizes with much higher dimensions of 3D medical images, the lack of hierarchical design for MAE may hinder the performance of downstream tasks. In this paper, we propose a novel Mask in Mask (MiM) pre-training framework for 3D medical images, which aims to advance MAE by learning discriminative representation from hierarchical visual tokens across varying scales. We introduce multiple levels of granularity for masked inputs from the volume, which are then reconstructed simultaneously ranging at both fine and coarse levels. Additionally, a cross-level alignment mechanism is applied to adjacent level volumes to enforce anatomical similarity hierarchically. Furthermore, we adopt a hybrid backbone to enhance the hierarchical representation learning efficiently during the pre-training. MiM was pre-trained on a large scale of available 3D volumetric images, i.e., Computed Tomography (CT) images containing various body parts. Extensive experiments on twelve public datasets demonstrate the superiority of MiM over other SSL methods in organ/tumor segmentation and disease classification. We further scale up the MiM to large pre-training datasets with more than 10k volumes, showing that large-scale pre-training can further enhance the performance of downstream tasks. © 1982-2012 IEEE.

关键词： Image segmentation

来源：评论

学校读者我要写书评

暂无评论

Accommodating Self-attentional Heterophily Topology into High- and Low-pass Graph Convolutional Network for Skeleton-based Action Recognition

Accommodating Self-attentional Heterophily Topology into Hig...

引用

International Joint Conference on Neural Networks (IJCNN)

作者： Chao Wei Zhidong Deng Department of Computer Science and Technology Beijing National Research Center for Information Science and Technology (BNRist) State Key Laboratory of Intelligent Technology and Systems Institute for Artificial Intelligence at Tsinghua University (THUAI) Tsinghua University Beijing China

In the scene of human skeleton-based action recognition, graph convolutional network (GCN) is widely applied to model human action dynamics and achieves extraordinary results, where graph topology is in charge of feature extraction and plays a crucial role in learning representation in GCNs. This paper presents a novel high- and low-pass graph convolutional network (HLP-GCN), aiming to effectively aggregate joint features based on learned homophily and heterophily topologies using self-attention modules for skeleton-based action recognition tasks. Specifically, such an HLP-GCN framework first learns homophily and heterophily topologies dynamically so as to capture different correlations between input samples and then combine similar information arising from homophily topologies with dissimilar ones from heterophily topologies to get the final embeddings. Furthermore, a temporal modeling module is added to this HLPGCN framework, and a new pooling strategy called global node (GNode) is cascaded subsequently, in order to have the extraction of more robust spatiotemporal features for skeleton action recognition. Finally, the experimental results demonstrate that our method outperforms previous state-of-the-art methods by 0.2%, 0.4%, and 0.3% on three large-scale action recognition datasets including NTU RGB+D X-Set, NTU RGB+D 120 X-Sub, and NW-UCLA, respectively.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Unified Question Generation with Continual Lifelong Learning

arXiv

引用

arXiv 2022年

作者： Yuan, Wei Yin, Hongzhi He, Tieke Chen, Tong Wang, Qiufeng Cui, Lizhen School of Information Technology and Electrical Engineering The University of Queensland BrisbaneQLD Australia State Key Laboratory for Novel Software Technology Nanjing University Nanjing China Department of School of Computer Science and Engineering Southeast University Nanjing China School of Software Shandong University Jinan China

Question Generation (QG), as a challenging Natural Language Processing task, aims at generating questions based on given answers and context. Existing QG methods mainly focus on building or training models for specific QG datasets. These works are subject to two major limitations: (1) They are dedicated to specific QG formats (e.g., answer-extraction or multi-choice QG), therefore, if we want to address a new format of QG, a re-design of the QG model is required. (2) Optimal performance is only achieved on the dataset they were just trained on. As a result, we have to train and keep various QG models for different QG datasets, which is resource-intensive and ungeneralizable. To solve the problems, we propose a model named Unified-QG based on lifelong learning techniques, which can continually learn QG tasks across different datasets and formats. Specifically, we first build a format-convert encoding to transform different kinds of QG formats into a unified representation. Then, a method named STRIDER (SimilariTy RegularIzed Difficult Example Replay) is built to alleviate catastrophic forgetting in continual QG learning. Extensive experiments were conducted on 8 QG datasets across 4 QG formats (answer-extraction, answer-abstraction, multi-choice, and boolean QG) to demonstrate the effectiveness of our approach. Experimental results demonstrate that our Unified-QG can effectively and continually adapt to QG tasks when datasets and formats vary. In addition, we verify the ability of a single trained Unified-QG model in improving 8 Question Answering (QA) systems' performance through generating synthetic QA data. © 2022, CC BY-NC-SA.

关键词： Extraction

来源：评论

学校读者我要写书评

暂无评论

Learnable Flow Model Conditioned on Graph Representation Memory for Anomaly Detection

Learnable Flow Model Conditioned on Graph Representation Mem...

引用

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Ziyu Zhu Wenlei Liu Zhidong Deng Department of Computer Science and Technology Tsinghua University Beijing China State Key Laboratory of Intelligent Technology and Systems Beijing National Research Center for Information Science and Technology (BNRist) Institute for Artificial Intelligence at Tsinghua University (THUAI)

Anomaly detection could be applied in a wide range of fields from industrial scene to medical imaging analysis. Although invertible flow models are developed to accomplish unsupervised anomaly detection, they are usually hard to train and have limited capabilities of accurately modeling the distribution of normal samples. To address this problem, we propose a novel enhanced flow model conditioned on graph representation memory (FlowGRM) for visual surface defect detection. In our FlowGRM, a graph neural network is trained to query graph embedding features from memory bank and incorporate them into a learnable flow model as conditional information. Such memorized conditional information, together with the invertible flow model, is jointly optimized to give rise to a probability density score. Experimental results obtained on the industrial visual anomaly detection datasets including MVTec and MTD show that the proposed model has state-of-the-art performance, which could even reach up to 99.6% and 99.2% AUROC scores on the above two bench-marks, respectively. Meanwhile, it requires fewer parameters and saves over 80% training time compared to existing flow models.

关键词： Training Visualization Time series analysis Signal processing Graph neural networks Acoustics Speech processing

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：