检索结果-内蒙古大学图书馆

Data science and computer Application (ICDSCA), IEEE International Conference on

作者： Kangjie Ning Chunpeng Zhou Zhi Yu School of Software Technology Zhejiang University Zhejiang Provincial Key Laboratory of Service Robot College of Computer Science Zhejiang University Hangzhou China School of Computer Science Zhejiang University Zhejiang Provincial Key Laboratory of Service Robot College of Computer Science Zhejiang University Hangzhou China

The detection of distresses with ground-penetrating radar (GPR) images is a highly challenging task due to their complex structure, blurred boundaries, and varying scales. The Segment Anything Model (SAM) has demonstrated impressive performance in natural image segmentation tasks with its zero-shot segmentation capability and high efficiency, which effectively reduces annotation time. However, the applicability of SAM for GPR image analysis remains till now. To this end, we construct a GPR image dataset comprising 395 samples with segmention, which are jointly confirmed and annotated by multiple domain experts. We comprehensively evaluate the performance of various mainstream deep learning image segmentation models and different SAM models on this dataset, and analyze the results in detail. Our experiments confirm that SAM's performance in detecting underground radar distresses is limited or even completely inadequate. These findings demonstrate that SAM's current zero-shot segmentation capability is insufficient for direct application in subgrade distress with GPR data.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Attention Beats Linear for Fast Implicit Neural Representation Generation

arXiv

引用

arXiv 2024年

作者： Zhang, Shuyi Liu, Ke Gu, Jingjun Cai, Xiaoxu Wang, Zhihua Bu, Jiajun Wang, Haishuai Zhejiang Provincial Key Laboratory of Service Robot College of Computer Science Zhejiang University Hangzhou China Shanghai Artificial Intelligence Laboratory Shanghai China

Implicit Neural Representation (INR) has gained increasing popularity as a data representation method, serving as a prerequisite for innovative generation models. Unlike gradient-based methods, which exhibit lower efficiency in inference, the adoption of hyper-network for generating parameters in Multi-Layer Perceptrons (MLP), responsible for executing INR functions, has surfaced as a promising and efficient alternative. However, as a global continuous function, MLP is challenging in modeling highly discontinuous signals, resulting in slow convergence during the training phase and inaccurate reconstruction performance. Moreover, MLP requires massive representation parameters, which implies inefficiencies in data representation. In this paper, we propose a novel Attention-based Localized INR (ANR) composed of a localized attention layer (LAL) and a global MLP that integrates coordinate features with data features and converts them to meaningful outputs. Subsequently, we design an instance representation framework that delivers a transformer-like hyper-network to represent data instances as a compact representation vector. With instance-specific representation vector and instance-agnostic ANR parameters, the target signals are well reconstructed as a continuous function. We further address aliasing artifacts with variational coordinates when obtaining the super-resolution inference results. Extensive experimentation across four datasets showcases the notable efficacy of our ANR method, e.g. enhancing the PSNR value from 37.95dB to 47.25dB on the CelebA dataset. Code is released at https://***/Roninton/ANR. © 2024, CC BY-NC-SA.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Efficient learning of robust quadruped bounding usingpretrained neural networks

引用

IET Cyber-Systems and robotics 2022年第4期4卷 331-338页

作者： Zhicheng Wang Anqiao Li Yixiao Zheng Anhuan Xie Zhibin Li Jun Wu Qiuguo Zhu Institute of Cyber-Systems and Control Zhejiang UniversityZhejiangChina Intelligent Robot Research Center Zhejiang LabZhejiangChina Department of Computer Science University College LondonLondonUK State Key Laboratory of Industrial Control Technology Zhejiang UniversityZhejiangChina

Bounding is one of the important gaits in quadrupedal locomotion for negotiating *** authors proposed an effective approach that can learn robust bounding gaits more efficiently despite its large variation in dynamic body *** authors first pretrained the neural network(NN)based on data from a robot operated by conventional model-based controllers,and then further optimised the pretrained NN via deep reinforcement learning(DRL).In particular,the authors designed a reward function considering contact points and phases to enforce the gait symmetry and periodicity,which improved the bounding *** NN-based feedback controller was learned in the simulation and directly deployed on the real quadruped robot Jueying Mini successfully.A variety of environments are presented both indoors and outdoors with the authors’*** authors’approach shows efficient computing and good locomotion results by the Jueying Mini quadrupedal robot bounding over uneven *** cover image is based on the Research Article Efficient learning of robust quadruped bounding using pretrained neural networks by Zhicheng Wang et al.,https://***/10.1049/csy2.12062.

关键词： legged locomotion reinforcement learning robot learning

来源：评论

学校读者我要写书评

暂无评论

An Artificial Intelligence-Based Segmentation Auxiliary Diagnostic System for Focal Liver Lesions

An Artificial Intelligence-Based Segmentation Auxiliary Diag...

引用

Mechatronics, IoT and Industrial Informatics (ICMIII), International Conference on

作者： Jingjun Gu Zhi Yu Wei Zhang Haoyu Hu Zhejiang Provincial Key Laboratory of Service Robot College of Computer Science Zhejiang University Hangzhou China School of Software Technology Zhejiang University Hangzhou China

ISBN: (数字)9798350386639

ISBN: (纸本)9798350386646

In this paper, we introduce an artificial intelligence-based system designed for the segmentation and auxiliary diagnosis of focal liver lesions. This system can effectively segment lesions in both singlephase non-contrast CT images and multiphase contrast-enhanced CT images. Additionally, it has the capability to synchronize CT data from Picture Archiving and Communication Systems (PACS) and to visualize the segmentation results. Our AI system aims to meet the demand for rapid screening of liver lesions in primary hospitals and physical examination institutions, while also fulfilling the more precise segmentation needs of provincial and municipal hospitals.

关键词： Image segmentation Hospitals Computed tomography Scalability Liver Data visualization Picture archiving and communication systems

来源：评论

学校读者我要写书评

暂无评论

Multi-label Classification of Road Cracks Based on Semantic-Aware Mixup

Multi-label Classification of Road Cracks Based on Semantic-...

引用

computers, Information Processing and Advanced Education (CIPAE), 2020 International Conference on

作者： Chengyang Xu Chunpeng Zhou Zhi Yu Zhejiang Provincial Key Laboratory of Service Robot School of Software Technology Zhejiang University Hangzhou China Zhejiang Provincial Key Laboratory of Service Robot College of Computer Science Zhejiang University Hangzhou China Zhejiang Provincial Key Laboratory of Service Robot School of Software Technology Zhejiang University Ningbo China

In the realm of road crack detection scenarios, Mixup, a widely used data augmentation technique, exhibits constrained effectiveness in multi-label classification, unlike its performance on general datasets. This paper investigates the reasons behind this limitation, identifying two main factors: (1) the mutually exclusive nature of road crack labels and (2) the potential generation of incorrect labels during image mixing. To address these issues, we propose an approach called Semantic-Aware Mixup (SA-Mixup), which focuses on enhancing crack samples specifically. Experimental evaluations on three distinct road crack datasets demonstrate the effectiveness and simplicity of our proposed method. The results highlight significant improvements in model performance, offering a promising solution for accurate and efficient multi-label classification in road crack detection tasks.

关键词：

来源：评论

学校读者我要写书评

暂无评论

CPDG: A Contrastive Pre-Training Method for Dynamic Graph Neural Networks

CPDG: A Contrastive Pre-Training Method for Dynamic Graph Ne...

引用

International Conference on Data Engineering

作者： Yuanchen Bei Hao Xu Sheng Zhou Huixuan Chi Haishuai Wang Mengdi Zhang Zhao Li Jiajun Bu Zhejiang Provincial Key Laboratory of Service Robot College of Computer Science and Technology Zhejiang University Meituan School of Software Technology Zhejiang University Institute of Computing Technology Chinese Academy of Science

ISBN: (数字)9798350317152

ISBN: (纸本)9798350317169

Dynamic graph data mining has gained popularity in recent years due to the rich information contained in dynamic graphs and their widespread use in the real world. Despite the advances in dynamic graph neural networks (DGNNs), the rich information and diverse downstream tasks have posed significant difficulties for the practical application of DGNNs in industrial scenarios. To this end, in this paper, we propose to address them by pre-training and present the Contrastive Pre-Training Method for Dynamic Graph Neural Networks (CPDG). CPDG tackles the challenges of pre-training for DGNNs, including generalization capability and long-short term modeling capability, through a flexible structural-temporal subgraph sampler along with structural-temporal contrastive pre-training schemes. Extensive experiments conducted on both large-scale research and industrial dynamic graph datasets show that CPDG outperforms existing methods in dynamic graph pre-training for various downstream tasks under three transfer settings.

关键词： Data engineering Graph neural networks Data mining Task analysis Recommender systems

来源：评论

学校读者我要写书评

暂无评论

LORE: Logical Location Regression Network for Table Structure Recognition

arXiv

引用

arXiv 2023年

作者： Xing, Hangdi Gao, Feiyu Long, Rujiao Bu, Jiajun Zheng, Qi Li, Liangcheng Yao, Cong Yu, Zhi Zhejiang Provincial Key Laboratory of Service Robot College of Computer Science Zhejiang University China Zhejiang Provincial Key Laboratory of Service Robot School of Software Technology Zhejiang University China DAMO Academy Alibaba Group Hangzhou China

Table structure recognition (TSR) aims at extracting tables in images into machine-understandable formats. Recent methods solve this problem by predicting the adjacency relations of detected cell boxes, or learning to generate the corresponding markup sequences from the table images. However, they either count on additional heuristic rules to recover the table structures, or require a huge amount of training data and time-consuming sequential decoders. In this paper, we propose an alternative paradigm. We model TSR as a logical location regression problem and propose a new TSR framework called LORE, standing for LOgical location REgression network, which for the first time combines logical location regression together with spatial location regression of table cells. Our proposed LORE is conceptually simpler, easier to train and more accurate than previous TSR models of other paradigms. Experiments on standard benchmarks demonstrate that LORE consistently outperforms prior arts. Code is available at https://***/AlibabaResearch/AdvancedLiterateMachinery/ tree/main/DocumentUnderstanding/LORE-TSR. Copyright © 2023, The Authors. All rights reserved.

关键词： Location

来源：评论

学校读者我要写书评

暂无评论

A Heatmap Regression Approach for Right Ventricle Anatomical Landmarks Localization on Echocardiography

A Heatmap Regression Approach for Right Ventricle Anatomical...

引用

Image Processing and computer Applications (ICIPCA), IEEE International Conference on

作者： Hailiang Wan Jingjun Gu Zhi Yu Shuyi Song Polytechnic Institute Zhejiang University Hangzhou China Zhejiang Provincial Key Laboratory of Service Robot College of Computer Science Zhejiang University Hangzhou China Zhejiang Provincial Key Laboratory of Service Robot School of Software Technology Zhejiang University Ningbo China DBAPPSecurity Ltd. College of Computer Science Zhejiang University Hangzhou China

In the process of quantitative cardiac analysis based on echocardiography, the localization of right ventricle anatomical landmarks is a critical step. However, manual labeling of these landmarks by professional physicians is a costly, time-consuming, and labor-intensive process that is also prone to both inter-observer and intra-observer variations. Therefore, we propose n heatmap regression approach to automatically locate right ventricle anatomical landmarks on echocardiography. Firstly, we propose an improved U-Net model, which has a stronger ability to extract features and exploit relationships between the coordinates of landmarks. Secondly, we analyze the limitations of L2 loss in our task. Building on this analysis, we propose a new loss function, Weighted Smooth L1 loss, which not only balances foreground and background pixels but also makes the model more robust to noise. Experiments show that the proposed improved U-Net and Weighted Smooth L1 loss can effectively enhance the landmark localization effect of the model.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Rethinking Propagation for Unsupervised Graph Domain Adaptation

arXiv

引用

arXiv 2024年

作者： Liu, Meihan Fang, Zeyu Zhang, Zhen Gu, Ming Zhou, Sheng Wang, Xin Bu, Jiajun Zhejiang Provincial Key Laboratory of Service Robot College of Computer Science Zhejiang University China School of Software Technology Zhejiang University China Department of Computer Science National University of Singapore Singapore Department of Computer Science and Technology Tsinghua University China

Unsupervised Graph Domain Adaptation (UGDA) aims to transfer knowledge from a labelled source graph to an unlabelled target graph in order to address the distribution shifts between graph domains. Previous works have primarily focused on aligning data from the source and target graph in the representation space learned by graph neural networks (GNNs). However, the inherent generalization capability of GNNs has been largely overlooked. Motivated by our empirical analysis, we reevaluate the role of GNNs in graph domain adaptation and uncover the pivotal role of the propagation process in GNNs for adapting to different graph domains. We provide a comprehensive theoretical analysis of UGDA and derive a generalization bound for multi-layer GNNs. By formulating GNN Lipschitz for k-layer GNNs, we show that the target risk bound can be tighter by removing propagation layers in source graph and stacking multiple propagation layers in target graph. Based on the empirical and theoretical analysis mentioned above, we propose a simple yet effective approach called A2GNN for graph domain adaptation. Through extensive experiments on real-world datasets, we demonstrate the effectiveness of our proposed A2GNN framework. © 2024, CC BY.

关键词： Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

Road Surface Damage Detection Based on Knowledge Distillation

Road Surface Damage Detection Based on Knowledge Distillatio...

引用

Image Processing and computer Applications (ICIPCA), IEEE International Conference on

作者： Qiao Sun Yi Yang Jingjun Gu Zhi Yu Zhejiang Provincial Key Laboratory of Service Robot College of Computer Science Zhejiang University Hangzhou China School of Electronic Engineering Naval University of Engineering Wuhan China Zhejiang Provincial Key Laboratory of Service Robot School of Software Technology Zhejiang University Ningbo China

As an important public infrastructure, road contributes to economic development and growth, brings significant social benefits. However, due to a variety of factors, the road surfaces will wear and deteriorate over time. When there is a slight crack in the pavement, most will deteriorate rapidly if not repaired promptly, which can greatly affect the life of the road. Therefore, developing a road surface damage detection algorithm is an urgent demand. We use the knowledge distillation technology to solve this problem. First, the similarity-preserving knowledge distillation was improved by adding a mask block to make the model transmit only the key feature information of the local region. Secondly, by making the boundary box of the teacher output and ground truth do non-maximum suppression, the image area of the distillation in the output layer is restricted, and the transmission of background information by the teacher is reduced. the performance of the student is improved by 1.2% in mAP@0.5 on average and 2.2% in mAP@0.5:0.95 on average compared with the state-of-the-art method.

关键词：

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：