检索结果-内蒙古大学图书馆

SSRN 2024年

作者： Zhang, Xiongtao Yin, Zezong Jiang, Yunliang Jiang, Yizhang Sun, Danfeng Liu, Yong School of Information Engineering Huzhou University Huzhou31300 China Zhejiang Key Laboratory of Intelligent Education Technology and Application Zhejiang Normal University Jinhua321004 China School of Computer Science and Technology Zhejiang Normal University Jinhua321004 China School of Artificial Intelligence and Computer Science Jiangnan University Wuxi 214122 China Key Laboratory of Image Processing and Intelligent Control Huazhong University of Science and Technology Ministry of Education Wuhan430074 China School of Computer Science Hangzhou Dianzi University Hangzhou310018 China College of Control Science and Engineering Zhejiang University Hangzhou310027 China

In order to enpower the hight performance as well as interpretability of low-order TSK fuzzy classifier, a born-again TSK fuzzy classier embedded with decoupled fuzzy dark knowledge distillation called HTSK-LLM-DKD is proposed in this study. HTSK-LLM DKD achieves the following distinctive characteristics: 1) It takes high-order TSK as teacher model and low-order TSK fuzzy classifier as student model, and leverages the proposed Least Learning Machine based Decoupling Knowledge Distillation to distill the fuzzy dark knowledge from high-order TSK fuzzy classifier to low-order TSK fuzzy classifier, which resulting in low-order TSK fuzzy classifier endowed with enhanced performance surpassing or at least comparable to high-order TSK fuzzy classifier;specifically 2) The Negative Euclidean distance between the output of teacher model and each class is employed to obtain the teacher logits, and then it computes teacher/student soft labels by the softmax function with distilling temperature parameter;3) By reformulating the Kullback Leibler divergence, it decouples fuzzy dark knowledge into target class knowledge and non-target class knowledge, and transfers them to student model, which taking HTSK-LLM-DKD with outstanding performance and relatively fast running speed. Extensive experiment results on UCI benchmarking datasets and real-world dataset Cleveland Heart Disease demonstrate the effectiveness of the proposed HTSK-LLM-DKD. © 2024, The Authors. All rights reserved.

关键词： Students

来源：评论

学校读者我要写书评

暂无评论

Modeling Inter-Intra Heterogeneity for Graph Federated Learning 39

Modeling Inter-Intra Heterogeneity for Graph Federated Learn...

引用

39th Annual AAAI Conference on Artificial intelligence, AAAI 2025

作者： Yu, Wentao Chen, Shuo Tong, Yongxin Gu, Tianlong Gong, Chen School of Computer Science and Engineering Nanjing University of Science and Technology China Center for Advanced Intelligence Project RIKEN Japan State Key Laboratory of Complex & Critical Software Environment Beihang University China Engineering Research Center of Trustworthy AI Ministry of Education Jinan University China Department of Automation Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University China

ISBN: (纸本)157735897X

Heterogeneity is a fundamental and challenging issue in federated learning, especially for the graph data due to the complex relationships among the graph nodes. To deal with the heterogeneity, lots of existing methods perform the weighted federation based on their calculated similarities between pairwise clients (i.e., subgraphs). However, their inter-subgraph similarities estimated with the outputs of local models are less reliable, because the final outputs of local models may not comprehensively represent the real distribution of subgraph data. In addition, they ignore the critical intra-heterogeneity which usually exists within each subgraph itself. To address these issues, we propose a novel Federated learning method by integrally modeling the Inter-Intra Heterogeneity (FedIIH). For the inter-subgraph relationship, we propose a novel hierarchical variational model to infer the whole distribution of subgraph data in a multi-level form, so that we can accurately characterize the inter-subgraph similarities with the global perspective. For the intra-heterogeneity, we disentangle the subgraph into multiple latent factors and partition the model parameters into multiple parts, where each part corresponds to a single latent factor. Our FedIIH not only properly computes the distribution similarities between subgraphs, but also learns disentangled representations that are robust to irrelevant factors within subgraphs, so that it successfully considers the inter- and intra- heterogeneity simultaneously. Extensive experiments on six homophilic and five heterophilic graph datasets in both non-overlapping and overlapping settings demonstrate the effectiveness of our method when compared with nine state-of-the-art methods. Specifically, FedIIH averagely outperforms the second-best method by a large margin of 5.79% on all heterophilic datasets. Copyright © 2025, Association for the Advancement of Artificial intelligence (***). All rights reserved.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Exploiting Distilled Learning for Deep Siamese Tracking

Exploiting Distilled Learning for Deep Siamese Tracking

引用

International Conference on Pattern Recognition

作者： Chengxin Liu Zhiguo Cao Wei Li Yang Xiao Shuaiyuan Du Angfan Zhu Key Laboratory of Image Processing and Intelligent Control Ministry of Education School of Artificial Intelligence and Automation Huazhong University of Science and Technology Wuhan China Queen Mary University of London

Existing deep siamese trackers are typically built on off-the-shelf CNN models for feature learning, with the demand for huge power consumption and memory storage. This limits current deep siamese trackers to be carried on resource-constrained devices like mobile phones, given factor that such a deployment normally requires cost-effective considerations. In this work, we address this issue by presenting a novel Distilled Learning Framework(DLF) for siamese tracking, which aims at learning tracking model with efficiency and high accuracy. Specifically, we propose two simple yet effective knowledge distillation strategies, denote as point-wise distillation and pairwise distillation, which are designed for transferring knowledge from a more discriminative teacher tracker into a compact student tracker. In this way, cost-effective and high performance tracking could be achieved. Extensive experiments on several tracking benchmarks demonstrate the effectiveness of our proposed method.

关键词： Power demand Pipelines Memory management Benchmark testing Mobile handsets Pattern recognition

来源：评论

学校读者我要写书评

暂无评论

RLSAC: Reinforcement Learning enhanced Sample Consensus for End-to-End Robust Estimation

RLSAC: Reinforcement Learning enhanced Sample Consensus for ...

引用

International Conference on Computer Vision (ICCV)

作者： Chang Nie Guangming Wang Zhe Liu Luca Cavalli Marc Pollefeys Hesheng Wang Department of Automation Key Laboratory of System Control and Information Processing of Ministry of Education Shanghai Jiao Tong University Department of Computer Science ETH Zürich MoE Key Lab of Artificial Intelligence AI Institute Shanghai Jiao Tong University Microsoft Mixed Reality and AI Zürich Lab

Robust estimation is a crucial and still challenging task, which involves estimating model parameters in noisy environments. Although conventional sampling consensus-based algorithms sample several times to achieve robustness, these algorithms cannot use data features and historical information effectively. In this paper, we propose RLSAC, a novel Reinforcement Learning enhanced SAmple Consensus framework for end-to-end robust estimation. RLSAC employs a graph neural network to utilize both data and memory features to guide exploring directions for sampling the next minimum set. The feedback of downstream tasks serves as the reward for unsupervised training. Therefore, RL-SAC can avoid differentiating to learn the features and the feedback of downstream tasks for end-to-end robust estimation. In addition, RLSAC integrates a state transition module that encodes both data and memory features. Our experimental results demonstrate that RLSAC can learn from features to gradually explore a better hypothesis. Through analysis, it is apparent that RLSAC can be easily transferred to other sampling consensus-based robust estimation tasks. To the best of our knowledge, RLSAC is also the first method that uses reinforcement learning to sample consensus for end-to-end robust estimation. We release our codes at https://***/IRMVLab/RLSAC.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Reliability on deep learning models: A comprehensive observation 6

Reliability on deep learning models: A comprehensive observa...

引用

6th International Symposium on System and Software Reliability, ISSSR 2020

作者： Zhang, Yuhong Xiao, Chunjing School of Artificial Intelligence and Big Data Key Laboratory of Grain Information Processing and Control Ministry of Education Zhengzhou China Henan Key Laboratory of Big Data Analysis and Processing Henan University Kaifeng China

ISBN: (纸本)9780738104973

This paper provides a comprehensive observation to examine the reliability of deep learning (DL) models. First, we will briefly introduce the essential background and kernel techniques in deep learning, such as downsampling and nonlinear discontinuity. Each of them may have some relation to the reliability of DL models. Then we discuss the inherent structural flaws of deep learning and the risk of unreliability that can result from it. Subsequently, we discuss various ways of generating adversarial samples that affect the DL model's reliability and corresponding preventive measures. Finally, we complete this observation by identifying current challenges and future trends for research. © 2020 IEEE.

关键词： Reliability

来源：评论

学校读者我要写书评

暂无评论

RegFormer: An Efficient Projection-Aware Transformer Network for Large-Scale Point Cloud Registration

arXiv

引用

arXiv 2023年

作者： Liu, Jiuming Wang, Guangming Liu, Zhe Jiang, Chaokang Pollefeys, Marc Wang, Hesheng Department of Automation Key Laboratory of System Control and Information Processing of Ministry of Education Shanghai Jiao Tong University China MoE Key Lab of Artificial Intelligence AI Institute Shanghai Jiao Tong University China China University of Mining and Technology China ETH Zürich Switzerland Microsoft United States

Although point cloud registration has achieved remarkable advances in object-level and indoor scenes, large-scale registration methods are rarely explored. Challenges mainly arise from the huge point number, complex distribution, and outliers of outdoor LiDAR scans. In addition, most existing registration works generally adopt a two-stage paradigm: They first find correspondences by extracting discriminative local features and then leverage estimators (eg. RANSAC) to filter outliers, which are highly dependent on well-designed descriptors and post-processing choices. To address these problems, we propose an end-to-end transformer network (RegFormer) for large-scale point cloud alignment without any further post-processing. Specifically, a projection-aware hierarchical transformer is proposed to capture long-range dependencies and filter outliers by extracting point features globally. Our transformer has linear complexity, which guarantees high efficiency even for large-scale scenes. Furthermore, to effectively reduce mismatches, a bijective association transformer is designed for regressing the initial transformation. Extensive experiments on KITTI and NuScenes datasets demonstrate that our RegFormer achieves competitive performance in terms of both accuracy and efficiency. Codes are available at https://***/IRMVLab/RegFormer. Copyright © 2023, The Authors. All rights reserved.

关键词： Efficiency

来源：评论

学校读者我要写书评

暂无评论

A Shape-Supervised Feature Fusion U-Net for Tubular Structure Segmentation

SSRN

引用

SSRN 2024年

作者： Yue, Jinghua Jin, Shuo Wang, Siyuan Zeng, Jianping Shan, Siqiao Liu, Bo Jiang, Nan Zhou, Fugen Image Processing Center Beihang University Beijing China Hepatopancreatobiliary Center Beijing Tsinghua Changgung Hospital Key Laboratory of Digital Intelligence Hepatology Ministry of Education School of Clinical Medicine Tsinghua University Beijing China Research Unit of Precision Hepatobiliary Surgery Paradigm Chinese Academy of Medical Sciences Beijing China Institute for Precision Medicine Tsinghua University Beijing China

Accurate segmentation of tubular structures, such as blood vessels and bile ducts, is pivotal for clinical diagnosis and subsequent treatment. However, challenges arise from their thin elongated structures, complex morphologies, and complex backgrounds, making it difficult to extract detailed and shape features and consequently posing challenges for automated segmentation. This paper introduces a novel approach, the shape-supervised feature fusion U-Net (SFU-Net), to address the segmentation of tubular structures. The proposed network extends the U-Net architecture by integrating a feature fusion module, effectively fuses different semantic information in both the encoder and decoder paths to facilitate the acquisition of detailed geometric representations of tubular structures. Furthermore, a novel distance-based tubularity loss is designed to capture boundary information and supervise segmentation integrity by imposing shape constraints. We conducted comprehensive experiments on three tubular structure segmentation tasks in medical imaging, including dilated biliary tree, hepatic vessel, and pulmonary artery. The experimental results illustrate the superior performance of our proposed method compared to other state-of-the-art approaches, underscoring its effectiveness in medical image segmentation with tubular structures. © 2024, The Authors. All rights reserved.

关键词： Blood vessels

来源：评论

学校读者我要写书评

暂无评论

An advanced actor-critic algorithm for training video game ai 1st

An advanced actor-critic algorithm for training video game a...

引用

1st International Conference on Neural Computing for Advanced Applications, NCAA 2020

作者： Zha, Zhong Yi Tang, Xue Song Wang, Bo Key Laboratory of Ministry of Education for Image Processing and Intelligent Control Artificial Intelligence and Automation School Huazhong University of Science and Technology Wuhan430074 China College of Information and Science Donghua University Shanghai China

ISBN: (纸本)9789811576690

This paper presents an improved deep reinforcement learning (DRL) algorithm, namely Advanced Actor Critic (AAC), which is based on Actor Critic (AC) algorithm, to the video game Artificial intelligence (AI) training. The advantage distribution estimator, Normal Constraint (NC) function and exploration based on confidence are introduced to improve the AC algorithm, in order to achieve accurate value estimate, select continuous spatial action correctly, and explorate effectively. The aim is to improve the performance of conventional AC algorithm in a complex environment. A case study of video game StarCraft II mini-game AI training is employed. The results verify that the improved algorithm effectively improves the performance in terms of the convergence rate, maximum reward, average reward in every 100 episode, and time to reach a specific reward, etc. The analysis on how these modifications improve the performance is also given through interpretation of the feature layers in the mini-games. © Springer Nature Singapore Pte Ltd. 2020.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Discovering the nuclear localization signal universe through a deep learning model with interpretable attention units

Patterns

引用

Patterns 2025年第6期6卷

作者： Li, Yi-Fan Pan, Xiaoyong Shen, Hong-Bin Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University and Key Laboratory of System Control and Information Processing Ministry of Education of China Shanghai200240 China

We describe NLSExplorer, an interpretable approach for nuclear localization signal (NLS) prediction. By utilizing the extracted information on nuclear-specific sites from the protein language model to assist in NLS detection, NLSExplorer achieves superior performance with greater than 10% improvement in the F1 score compared with existing methods on benchmark datasets and highlights other nuclear transport segments. We applied NLSExplorer to the nucleus-localized proteins in the Swiss-Prot database to extract valuable segments. A comprehensive analysis of these segments revealed a potential NLS landscape and uncovered features of nuclear transport segments across 416 species. This study introduces a powerful tool for exploring the NLS universe and provides a versatile network that can efficiently detect characteristic domains and motifs. © 2025 The Author(s)

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Memristor-based multilayer neural network with edge learning and weight update

引用

Neurocomputing 2025年 647卷

作者： Ningye Jiang Jupeng Xie Mingxuan Jiang Haoen Huang Zhigang Zeng School of Artificial Intelligence and Automation Huazhong University of Science and Technology Wuhan 430074 China Key Laboratory of Image Information Processing and Intelligent Control Ministry of Education of China Wuhan 430074 China

Memristor crossbar array (MCA) is a promising computing-in-memory architecture for accelerating vector-matrix multiplication (VMM). However, existing memristor-based circuits primarily perform forward computations without local learning capabilities. To address the issue, a memristor-based multilayer neural network (MNN) with edge learning is proposed, enabling error backpropagation and one-step weight update. In particular, the overall circuit design of a three-layer MNN is outlined, comprising three operational stages: forward computation, backpropagation, and weight update. Subsequently, the schematics of four adopted circuit blocks are detailed. By integrating the memory and weight update blocks with MCA, the proposed three-layer memristor-based MNN exhibits scalability and universality, allowing it to be extended to the deep neural network (DNN) architecture. Circuit simulations validate the effectiveness of the proposed scheme, while offline-online co-simulations demonstrate its strength in compensating for variations and non-ideal characteristics of physical devices. Comparative results further highlight the advantages of the proposed scheme in realizing full-analog edge learning circuits and performing error backpropagation processes efficiently.

关键词： Analog circuit design Computing-in-memory Edge learning Memristor crossbar array (MCA) Multilayer neural network

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：