检索结果-内蒙古大学图书馆

37th Conference on Neural Information Processing Systems, NeurIPS 2023

作者： Li, Zhixun Wang, Liang Sun, Xin Luo, Yifan Zhu, Yanqiao Chen, Dingshuo Luo, Yingtao Zhou, Xiangxin Liu, Qiang Wu, Shu Yu, Jeffrey Xu Department of Systems Engineering and Engineering Management The Chinese University of Hong Kong Hong Kong Center for Research on Intelligent Perception and Computing State Key Laboratory of Multimodal Artificial Intelligence Systems Institute of Automation Chinese Academy of Sciences China School of Artificial Intelligence University of Chinese Academy of Sciences China Department of Automation University of Science and Technology of China China School of Cyberspace Security Beijing University of Posts and Telecommunications China Department of Computer Science University of California Los Angeles United States Heinz College of Information Systems and Public Policy Machine Learning Department School of Computer Science Carnegie Mellon University United States

ISBN: (纸本)9781713899921

Graph Structure Learning (GSL) has recently garnered considerable attention due to its ability to optimize both the parameters of Graph Neural Networks (GNNs) and the computation graph structure simultaneously. Despite the proliferation of GSL methods developed in recent years, there is no standard experimental setting or fair comparison for performance evaluation, which creates a great obstacle to understanding the progress in this field. To fill this gap, we systematically analyze the performance of GSL in different scenarios and develop a comprehensive Graph Structure Learning Benchmark (GSLB) curated from 20 diverse graph datasets and 16 distinct GSL algorithms. Specifically, GSLB systematically investigates the characteristics of GSL in terms of three dimensions: effectiveness, robustness, and complexity. We comprehensively evaluate state-of-the-art GSL algorithms in node- and graph-level tasks, and analyze their performance in robust learning and model complexity. Further, to facilitate reproducible research, we have developed an easy-to-use library for training, evaluating, and visualizing different GSL methods. Empirical results of our extensive experiments demonstrate the ability of GSL and reveal its potential benefits on various downstream tasks, offering insights and opportunities for future research. The code of GSLB is available at: https://***/GSL-Benchmark/GSLB. © 2023 Neural information processing systems foundation. All rights reserved.

关键词： Graphic methods

来源：评论

学校读者我要写书评

暂无评论

Terrain recognition and gait cycle prediction using IMU

Terrain recognition and gait cycle prediction using IMU

引用

2021 IEEE International Conference on Real-Time computing and Robotics, RCAR 2021

作者： Wang, Zhuo Zhang, Yu Ni, Jiangpeng Wu, Xinyu Liu, Yida Ye, Xin Chen, Chunjie Shenzhen Institute of Advanced Technology Chinese Academy of Sciences Shenzhen China CAS Key Laboratory of Human-Machine Intelligence-Synergy Systems Shenzhen Institute of Advanced Technology Shenzhen China Guangdong Provincial Key Lab of Robotics and Intelligent System Shenzhen Institute of Advanced Technology Chinese Academy of Sciences Shenzhen China SZ DJI Technology Co. Ltd. Shenzhen China

ISBN: (纸本)9781665436786

It is well known that terrain recognition and gait cycle prediction are important for powered exoskeleton. However, only a few works have focused on the concerns of complexity of the control system caused by using redundant sensors. In this paper, only two IMU sensors are applied to collect information of the angle and angular velocity of the hip joint in the situation of level-ground walking, ramp ascent, and ramp descent. Based on information acquired from these two IMU sensors, two methods are proposed to achieve terrain recognition. One method uses the angle of the hip joint when the two legs intersect as the threshold of terrain recognition. It can identify the terrain (level-ground walking, ramp ascent, ramp descent) during stable walking, but it cannot recognize the transitional terrain (from level-ground walking to ramp ascent, from ramp ascent to ramp descent, and so on) and its robustness is limited. The other method selects the angle and angular velocity of the hip joints as the eigenvector, and uses SVM for terrain recognition. The accuracy of terrain recognition is improved from 69.7% to 100% after introducing the Gaussian kernel function instead of Linear kernel function. For gait cycle prediction, Wiener one step prediction is applied in predicting the GC. Compared to actual GC, the error from predicted GC based on mean prediction is more than 8.0%, while the error from Wiener on step prediction is less than 4.35%. © 2021 IEEE.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

Towards Disease-Aware Self-Supervised Dynamic Brain Network Learning For Mental Diagnosis

Towards Disease-Aware Self-Supervised Dynamic Brain Network ...

引用

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

作者： Zhiyong Jin Guangqi Wen Peng Cao Lingwen Liu Jinzhu Yang Xinrong Zhu Osmar R. Zaiane Fei Wang Computer Science and Engineering Northeastern University Shenyang China Key Laboratory of Intelligent Computing in Medical Image of Ministry of Education Northeastern University Shenyang China Department of Psychiatry Affiliated Nanjing Brain Hospital Early Intervention Unit Nanjing Medical University Nanjing China Alberta Machine Intelligence Institute University of Alberta Edmonton Alberta Canada

The dynamic brain network learning methods ignored the separation of redundant disease-irrelevant information, resulting in the model only achieving suboptimal diagnosis results. Meanwhile, the supervised learning scheme inevitably suffers from poor generalization due to the limited data. To address these problems, we propose a Self-supervised Dynamic Brain network Disentangled representation learning framework named SDBD, which incorporates 1) a dynamic topology-aware encoder for capturing diverse topological information, 2) a cross decoder for reconstructing the graph structure and 3) a spatio-temporal learning model based on the multi-head self-attention mechanism for classification. To disentangle the disease-related information from the dynamic brain networks, we design a temporal contrastive loss and a structure reconstruction loss. We evaluate our model on three real-world mental diseases: Autism Spectrum Disorder (ASD), Major Depressive Disorder (MDD), and Bipolar Disorder (BD). The results indicate significant improvements in our SDBD over the state-of-the-art methods owing to the disentangled disease-related information. Moreover, our method can identify the biomarkers associated with the diseases, which is consistent with the previous studies. To the best of our knowledge, our work is the first attempt to disentangle the disease-related information for the dynamic brain network analysis. The code is available at https://***/IntelliDAL/Graph/tree/main/SDBD.

关键词：

来源：评论

学校读者我要写书评

暂无评论

An Automatic Method for Brain Tumors Segmentation Based on Deep Convolutional Neural Network

An Automatic Method for Brain Tumors Segmentation Based on D...

引用

2021 IEEE International Conference on Medical Imaging Physics and Engineering, ICMIPE 2021

作者： Lin, Hui Zhang, Benjian Guo, Xin Guo, Dong Jing, Jia Huiyuan, Luo Yumeng, Zhao Hu, Min Ren, Fuji School of Physics Hefei University of Technology Hefei China School of Computer and Information Anhui Province Key Laboratory of Affective Computing and Advanced Intelligent Machine Hefei China School of Computer and Information Hefei University of Technology Hefei China Graduate School of Advanced Technology and Science University of Tokushima Tokushima Japan

ISBN: (纸本)9781665426084

[Purpose] Accurate outline of tumor targets is critical to a high quality radiotherapy plan. Manual segmentation is of great workload and has a strong artificial subjectivity. Using deep learning method to assist automatically segmenting of tumor target is the penetration and application of artificial intelligence in medicine. [Methods] A 6-layer model of deep Convolution Neural Network (CNN) has been constructed by taking advantage of different types of layers for brain tumor segmentation. This model is a 6 layer CNN model (6-CNN) composed of three convolution layers, two pool layers and one full connection layer. To obtain enough samples for 6-CNN model training, a patch-based technology has been adopted. That is to successively extract a local area from the whole image as a patch. And the center pixel value is taken as the pixel value of the whole patch. Similarly, the label of the center pixel is also taken as the label of the whole patch. Thus the 6-CNN model transforms the brain tumor image segmentation into patch classification based on the excellent classification characteristics of deep convolution neural network. The model combines the local features of patch, the information extracted from shallow network and the global features to predict the category label of the central pixel of patch. [Results] The model is validated on BRATS 2015 dataset and results show that the segmentation accuracy can be up to Dice Similarity Coefficient (DSC) 90\%\pm 4\%. [Conclusions] An automatic deep CNN segmentation model for brain tumors has been constructed based on MRI image patches, which is expected to assist or even substitute the manual segmentation of brain tumors. © 2021 IEEE.

关键词： Neural network models

来源：评论

学校读者我要写书评

暂无评论

Rethinking and Improving Relative Position Encoding for Vision Transformer

Rethinking and Improving Relative Position Encoding for Visi...

引用

International Conference on Computer Vision (ICCV)

作者： Kan Wu Houwen Peng Minghao Chen Jianlong Fu Hongyang Chao School of Computer Science and Engineering Sun Yat-sen University Ministry of Education The Key Laboratory of Machine Intelligence and Advanced Computing (Sun Yat-sen University) Microsoft Research Asia

ISBN: (纸本)9781665428132

Relative position encoding (RPE) is important for transformer to capture sequence ordering of input tokens. General efficacy has been proven in natural language processing. However, in computer vision, its efficacy is not well studied and even remains controversial, e.g., whether relative position encoding can work equally well as absolute position? In order to clarify this, we first review existing relative position encoding methods and analyze their pros and cons when applied in vision transformers. We then propose new relative position encoding methods dedicated to 2D images, called image RPE (iRPE). Our methods consider directional relative distance modeling as well as the interactions between queries and relative position embeddings in self-attention mechanism. The proposed iRPE methods are simple and lightweight. They can be easily plugged into transformer blocks. Experiments demonstrate that solely due to the proposed encoding methods, DeiT [21] and DETR [1] obtain up to 1.5% (top-1 Acc) and 1.3% (mAP) stable improvements over their original versions on ImageNet and COCO respectively, without tuning any extra hyperparameters such as learning rate and weight decay. Our ablation and analysis also yield interesting findings, some of which run counter to previous understanding. Code and models are open-sourced at https://***/microsoft/Cream/tree/main/iRPE.

关键词： Computer vision Image coding Codes Computational modeling Transformers Encoding Natural language processing

来源：评论

学校读者我要写书评

暂无评论

BGTplanner: Maximizing Training Accuracy for Differentially Private Federated Recommenders via Strategic Privacy Budget Allocation

arXiv

引用

arXiv 2024年

作者： Zhang, Xianzhi Zhou, Yipeng Hu, Miao Wu, Di Liao, Pengshan Guizani, Mohsen Sheng, Michael The School of Computer Science and Engineering Sun Yat-sen University Guangzhou China Guangdong Key Laboratory of Big Data Analysis and Processing Guangzhou China The Department of Computing Faculty of Science and Engineering Macquarie University Sydney Australia The Machine Learning Department Mohamed Bin Zayed University of Artificial Intelligence Abu Dhabi United Arab Emirates

To mitigate the rising concern about privacy leakage, the federated recommender (FR) paradigm emerges, in which decentralized clients co-train the recommendation model without exposing their raw user-item rating data. The differentially private federated recommender (DPFR) further enhances FR by injecting differentially private (DP) noises into clients. Yet, current DPFRs, suffering from noise distortion, cannot achieve satisfactory accuracy. Various efforts have been dedicated to improving DPFRs by adaptively allocating the privacy budget over the learning process. However, due to the intricate relation between privacy budget allocation and model accuracy, existing works are still far from maximizing DPFR accuracy. To address this challenge, we develop BGTplanner (Budget Planner) to strategically allocate the privacy budget for each round of DPFR training, improving overall training performance. Specifically, we leverage the Gaussian process regression and historical information to predict the change in recommendation accuracy with a certain allocated privacy budget. Additionally, Contextual Multi-Armed Bandit (CMAB) is harnessed to make privacy budget allocation decisions by reconciling the current improvement and long-term privacy constraints. Our extensive experimental results on real datasets demonstrate that BGTplanner achieves an average improvement of 6.76% in training performance compared to state-of-the-art baselines. Copyright © 2024, The Authors. All rights reserved.

关键词： Differential privacy

来源：评论

学校读者我要写书评

暂无评论

Multiple time scale motion images for action recognition 22

Multiple time scale motion images for action recognition

引用

22nd IEEE International Conference on E-Health Networking, Application and Services, HEALTHCOM 2020

作者： Cheng, Qin Ren, Ziliang Liu, Jianming Cheng, Jun School of Electronic Engineering and Automation Guilin University of Electronic Technology Guilin China Key Laboratory of Human-Machine Intelligence-Synergy Systems Shenzhen Institutes of Advanced Technology Chinese Academy of Sciences Shenzhen China The Chinese University of Hong Kong Hong Kong

ISBN: (纸本)9781728162676

This paper proposes a simple and effective approach for RGB-based action recognition using multiple streams Convolutional Neural Networks (ConvNets). We utilize a new method to represent temporal structure of RGB videos, named Motion Image (MI), which is constructed from the difference between frames with a certain time scale. Considering the different duration of different actions, multiple time scale sampling MIs can obtain more temporal information. Furthermore, we adopt multiple streams ConvNets, including MIs and RGB streams, to learn spatial-temporal features for action recognition. Our approach has been evaluated on UCF-101 and HMDB-51, and the experimental results demonstrate the effectiveness and significantly improve action recognition rate at a small computational cost. © 2021 IEEE.

关键词： Motion estimation

来源：评论

学校读者我要写书评

暂无评论

An Approach for EEG Data Augmentation Based on Deep Convolutional Generative Adversarial Network

An Approach for EEG Data Augmentation Based on Deep Convolut...

引用

IEEE International Conference on Cyborg and Bionic Systems (CBS)

作者： Yuanzhe Dong Xi Tang Fangning Tan Qingge Li Yingying Wang Huanqing Zhang Jun Xie Wenyuan Liang Guanglin Li Peng Fang CAS Key Laboratory of Human-Machine Intelligence-Synergy Systems Shenzhen Engineering Laboratory of Neural Rehabilitation Technology Shenzhen Institute of Advanced Technology Shenzhen China Shenzhen College of Advanced Technology University of Chinese Academy of Sciences Shenzhen China School of Mechanical Engineering Xi‘an Jiaotong University Xi‘an China National Research Center for Rehabilitation Technical Aids Beijing China

High-quality Electroencephalogram (EEG) signals are crucial for BCI applications. In practice, processing and analysis of EEG are often hindered due to either data scarcity or data imbalance, which would lower the performance of EEG-based classification. In this work, we implemented a data augmentation approach based on Deep Convolutional Generative Adversarial Network (DCGAN), which employed a generator and a discriminator to compete with each other to gain high-quality data for augmentation. In the approach, one-dimensional (ID) convolution layers were used by the discriminator to extract time features of raw signals, and ID transposed convolutional layers were adopted by the generator to synthesize new data. The quality of synthesized data was analyzed in time domain, and as well in both frequency and time-frequency domains by using Fast Fourier Transform and Continuous Wavelet Transform. The experimental results showed that the synthesized data have very similar temporal and spectral features to the raw data, which confirmed the effectiveness of our proposed method in synthesizing useable data from limited original signals with a concise structure and fewer network parameters. This work provides a possible deep learning method for EEG data augmentation with a lot of application potentials in BCI areas.

关键词： Time-frequency analysis Wavelet domain Convolution Generative adversarial networks Feature extraction Wavelet analysis Electroencephalography

来源：评论

学校读者我要写书评

暂无评论

A Compact Dead-Zone-Free Multi-Channel Stimulation and Recording System for Precision Peripheral Nerve Modulation

A Compact Dead-Zone-Free Multi-Channel Stimulation and Recor...

引用

Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)

作者： Yanjie Xing Weihuang Chen Danlei Yu Xiang Gao Jieting Bao Kedi Xu Department of Biomedical Engineering Qiushi Academy for Advanced Studies (QAAS) Zhejiang Zhejiang Provincial Key Laboratory of Cardio-Cerebral Vascular Detection Technology and Medicinal Effectiveness Appraisal Key Laboratory of Biomedical Engineering of Education Ministry Zhejiang University Hangzhou China Institute of Advanced Digital Technology and Instrument Zhejiang University Hangzhou China Zhejiang Provincial Key Laboratory for Network Multimedia Technologies Zhejiang University Hangzhou China Binjiang Institude of Zhejiang University Hangzhou China Nanhu Brain-Computer Interface Institute Hangzhou China Department of Biomedical Engineering Zhejiang Provincial Key Laboratory of Cardio-Cerebral Vascular Detection Technology and Medicinal Effectiveness Appraisal Key Laboratory of Biomedical Engineering of Education Ministry Zhejiang University Hangzhou China Qiushi Academy for Advanced Studies (QAAS) Zhejiang University Hangzhou China The State Key Lab of Brain-Machine Intelligence Zhejiang University Hangzhou China

ISBN: (数字)9798350371499

ISBN: (纸本)9798350371505

Peripheral nerve modulation has a certain potential in motor recovery, but the majority of systems suffer from the drawback of dead time caused by switching channels to achieve multi-channel stimulation. As a response to this challenge, we have developed a peripheral nerve modulation device with 64-channel recording and 16-channel stimulation without dead time. Through modulation of the rat sciatic nerve, we identified the crucial role of initial phase and amplitude in the outcomes of dual-channel combined stimulation. This design not only enhanced neural activation capabilities with constant power supply but also achieves more precise neural activation and reduces unintentional activation.

关键词： Phase modulation Power supplies Switches Motors Recording Engineering in medicine and biology

来源：评论

学校读者我要写书评

暂无评论

Reference-based defect detection network

arXiv

引用

arXiv 2021年

作者： Zeng, Zhaoyang Liu, Bei Fu, Jianlong Chao, Hongyang School of Computer Science and Engineering Sun Yat-sen University Key Laboratory of Machine Intelligence and Advanced Computing Sun Yat-sen University Ministry of Education Microsoft Research Asia

The defect detection task can be regarded as a realistic scenario of object detection in the computer vision field and it is widely used in the industrial field. Directly applying vanilla object detector to defect detection task can achieve promising results, while there still exists challenging issues that have not been solved. The first issue is the texture shift which means a trained defect detector model will be easily affected by unseen texture, and the second issue is partial visual confusion which indicates that a partial defect box is visually similar with a complete box. To tackle these two problems, we propose a Reference-based Defect Detection Network (RDDN). Specifically, we introduce template reference and context reference to against those two problems, respectively. Template reference can reduce the texture shift from image, feature or region levels, and encourage the detectors to focus more on the defective area as a result. We can use either well-aligned template images or the outputs of a pseudo template generator as template references in this work, and they are jointly trained with detectors by the supervision of normal samples. To solve the partial visual confusion issue, we propose to leverage the carried context information of context reference, which is the concentric bigger box of each region proposal, to perform more accurate region classification and regression. Experiments on two defect detection datasets demonstrate the effectiveness of our proposed approach. Copyright © 2021, The Authors. All rights reserved.

关键词： Defects

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：