检索结果-内蒙古大学图书馆

Combinatorial Learning of Robust Deep Graph Matching: An Embedding Based Approach

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE 2023年第6期45卷 6984-7000页

作者： Wang, Runzhong Yan, Junchi Yang, Xiaokang Shanghai Jiao Tong Univ AI Inst Dept Comp Sci & Engn MoE Key Lab Artificial Intelligence Shanghai 200240 Peoples R China Shanghai Jiao Tong Univ AI Inst MoE Key Lab Artificial Intelligence Shanghai 200240 Peoples R China

Graph matching aims to establish node correspondence between two graphs, which has been a fundamental problem for its NP-hard nature. One practical consideration is the effective modeling of the affinity function in the presence of noise, such that the mathematically optimal matching result is also physically meaningful. This paper resorts to deep neural networks to learn the node and edge feature, as well as the affinity model for graph matching in an end-to-end fashion. The learning is supervised by combinatorial permutation loss over nodes. Specifically, the parameters belong to convolutional neural networks for image feature extraction, graph neural networks for node embedding that convert the structural (beyond second-order) information into node-wise features that leads to a linear assignment problem, as well as the affinity kernel between two graphs. Our approach enjoys flexibility in that the permutation loss is agnostic to the number of nodes, and the embedding model is shared among nodes such that the network can deal with varying numbers of nodes for both training and inference. Moreover, our network is class-agnostic. Experimental results on extensive benchmarks show its state-of-the-art performance. It bears some generalization capability across categories and datasets, and is capable for robust matching against outliers.

关键词： Mathematical model Feature extraction Peer-to-peer computing Training Optimization Tensors Pattern matching Graph matching deep learning graph embedding combinatorial optimization

来源：评论

学校读者我要写书评

暂无评论

Predicting Stress-Strain Curve with Confidence: Balance Between Data Minimization and Uncertainty Quantification by a Dual Bayesian Model

引用

POLYMERS 2025年第4期17卷 550-550页

作者： Li, Tianyi Chen, Zhengyuan Zhang, Zhen Wei, Zhenhua Zhong, Gan-Ji Li, Zhong-Ming Liu, Han Sichuan Univ SOlids inFormaT AI Lab SOFT AI Lab Chengdu 610065 Peoples R China Sichuan Univ Coll Polymer Sci & Engn Chengdu 610065 Peoples R China Chengdu Univ Technol Coll Math & Phys Chengdu 610059 Peoples R China Southern Univ Sci & Technol Dept Ocean Sci & Engn Shenzhen 518055 Peoples R China Sichuan Univ State Key Lab Polymer Mat Engn Chengdu 610065 Peoples R China

Driven by polymer processing-property data, machine learning (ML) presents an efficient paradigm in predicting the stress-strain curve. However, it is generally challenged by (i) the deficiency of training data, (ii) the one-to-many issue of processing-property relationship (i.e., aleatoric uncertainty), and (iii) the unawareness of model uncertainty (i.e., epistemic uncertainty). Here, leveraging a Bayesian neural network (BNN) and a recently proposed dual-architected model for curve prediction, we introduce a dual Bayesian model that enables accurate prediction of the stress-strain curve while distinguishing between aleatoric and epistemic uncertainty at each processing condition. The model is trained using a Taguchi array dataset that minimizes the data size while maximizing the representativeness of 27 samples in a 4D processing parameter space, significantly reducing data requirements. By incorporating hidden layers and output-distribution layers, the model quantifies both aleatoric and epistemic uncertainty, aligning with experimental data fluctuations, and provides a 95% confidence interval for stress-strain predictions at each processing condition. Overall, this study establishes an uncertainty-aware framework for curve property prediction with reliable, modest uncertainty at a small data size, thus balancing data minimization and uncertainty quantification.

关键词： polymeric materials mechanical behavior machine learning prediction uncertainty polymer processing

来源：评论

学校读者我要写书评

暂无评论

DeepFlame 2.0: A new version for fully GPU-native machine learning accelerated reacting flow simulations under low-Mach conditions

引用

COMPUTER PHYSICS COMMUNICATIONS 2025年 312卷

作者： Mao, Runze Dong, Xinyu Bai, Xuan Wu, Ziheng Dang, Guanlin Li, Han Chen, Zhi X. Peking Univ Coll Engn State Key Lab Turbulence & Complex Syst Beijing 100871 Peoples R China AI Sci Inst Beijing 100080 Peoples R China

This paper presents DeepFlame v2.0, a significant computational framework upgrade designed for highperformance combustion simulations on GPU-based heterogeneous architectures. The updated version implements a comprehensive CUDA-accelerated architecture incorporating fundamental combustion modelling components, including: implicit/explicit finite volume method (FVM) discretisation schemes, chemical kinetics integrators, thermophysical property models, and subgrid-scale closures for both fluid dynamics and combustion processes. The redesigned code supports diverse boundary conditions and discretisation schemes for broad applicability across combustion configurations. key performance optimisations integrate advanced CUDA features including data coalescing techniques, CUDA Graphs for kernel scheduling, and NCCL-based multi-GPU communication. Validation studies employing the fully-implicit low-Mach solver demonstrate two-order-of-magnitude acceleration compared to conventional CPU implementations across canonical test cases, while maintaining numerical accuracy.

关键词： Computational fluid dynamic Compressible reacting flow GPU acceleration High performance computing Machine learning

来源：评论

学校读者我要写书评

暂无评论

ProSide: Knowledge Projector and Sideway for Pre-trained Language Models 13th

ProSide: Knowledge Projector and Sideway for Pre-trained Lan...

引用

13th International Conference on Natural Language Processing and Chinese Computing

作者： He, Chaofan Lu, Gewei Shen, Liping Shanghai Jiao Tong Univ AI Inst Dept Comp Sci & Engn X LANCE LabMoE Key Lab Artificial Intelligence Shanghai Peoples R China

ISBN: (纸本)9789819794331;9789819794348

Research indicates that incorporating external knowledge into pre-trained language models (PLMs) can enhance their performance on knowledge-driven downstream tasks. However, most approaches either require retraining the model or hardly maintain complete information of knowledge graphs (KGs). In this paper, we introduce a simple but effective ProSide module for PLMs, which includes two components: Knowledge Projector and Knowledge Sideway. Knowledge Projector transforms knowledge representation from entity embedding in KG space to semantic space, while Knowledge Sideway retains complete KG information in sideway modules through pre-training. Our ProSide can accommodate different frozen language models, without the need for retraining them. Results on three common knowledge-driven tasks demonstrate that our ProSide method enhances model performance and reaches state-of-the-art level. Additionally, we provide further analysis and case studies to illustrate the mechanism from the perspective of representation space. The code and models are publicly available at Github (Code and models are publicly available at https://***/hcffffff/ProSide).

关键词： Knowledge Graph Pre-trained Language Models Parameter Efficient Fine-tuning Method

来源：评论

学校读者我要写书评

暂无评论

Soft neural interface with color adjusted PDMS encapsulation layer for spinal cord stimulation

引用

JOURNAL OF NEUROSCIENCE METHODS 2025年 417卷 110402页

作者： Wang, Minjie Zhang, Yuan Wang, aiping Gan, Zhongxue Zhang, Lihua Kang, Xiaoyang Fudan Univ Acad Engn & TechnolLab Neural Interface & Brain C Inst Meta MedState Key Lab Med NeurobiolMinist E Inst AI & RobotShanghai Engn Res Ctr AI & RobotM Shanghai Peoples R China Fudan Univ Yiwu Res Inst Yiwu Peoples R China

Background: Spinal cord stimulation (SCS) plays a crucial role in treating various neurological diseases. Utilizing soft spinal cord electrodes in SCS allows for a better fit with the physiological structure of the spinal cord and reduces tissue damage. Polydimethylsiloxane (PDMS) has emerged as an ideal material for soft bioelectronics. However, micromachining soft PDMS bioelectronics devices with low thermal effects and high uniformity remains challenging. New method: Here, we demonstrated a fully laser-micromachined soft neural interface for SCS. The native and color adjusted PDMS with variable absorbance characteristics were investigated in laser processing. In addition, we systematically evaluated the impact of electrode sizes on the electrochemical performance of neural interface. By fitting the equivalent circuit model, the electrochemical process of neural interface was revealed and the performance of the electrode was evaluated. The biocompatibility of color adjusted PDMS was confirmed by cytotoxicity assays. Finally, we validated the neural interface in mice. Results: Color adjusted PDMS has good biocompatibility and can significantly reduce the damage caused by thermal effects, enhancing the electrochemical performance of bioelectronic devices. The soft neural interface with color adjusted PDMS encapsulation layer can activate the motor function safely. Comparison with existing methods: The fully laser-micromachined soft neural interface was proposed for the first time. Compared with existing methods, this method showed low thermal effects, high uniformity, and could be easily scaled up. Conclusions: The fully laser-micromachined soft neural interface device with color adjusted PDMS encapsulation layer shows great promise for applications in SCS.

关键词： Neural electrode Spinal cord stimulation Laser micro-machining Oily pigment PDMS Electrochemistry

来源：评论

学校读者我要写书评

暂无评论

Investigating water structure and dynamics at metal/ water interfaces from classical, ab initio to machine learning molecular dynamics

引用

CURRENT OPINION IN ELECTROCHEMISTRY 2025年 49卷

作者： Wang, Fei-Teng Cheng, Jun Xiamen Univ Coll Chem & Chem Engn State Key Lab Phys Chem Solid Surfaces iChem Xiamen 361005 Peoples R China IKKEM Lab AI Electrochem AI4EC Xiamen 361005 Peoples R China Xiamen Univ Inst Artificial Intelligence Xiamen 361005 Peoples R China

Metal-water interfaces are central to a wide range of crucial processes, including energy storage, energy conversion, and corrosion. Understanding the detailed structure and dynamics of water molecules at these interfaces is essential for unraveling the fundamental mechanisms driving these processes at the molecular level. Experimentally, a detection of interfacial structure and dynamics with high temporal and spatial resolution is lacking. The advances in machine learning molecular dynamics are offering an opportunity to address this issue with high accuracy and efficiency. To offer insights into the structure and dynamics, this review summarizes the progress made in determining the structure and dynamics of interfacial water molecules using molecular dynamics simulations. The possible application of machine learning molecular dynamics to address the fundamental challenges of simulating metal/water interfaces are also discussed.

关键词： metal/water interfaces water exchange dynamics structure and dy-namics

来源：评论

学校读者我要写书评

暂无评论

Generalizable Prompt Learning via Gradient Constrained Sharpness-Aware Minimization

引用

IEEE TRANSACTIONS ON MULTIMEDIA 2025年 27卷 1100-1113页

作者： Liu, Liangchen Wang, Nannan Zhou, Dawei Liu, Decheng Yang, Xi Gao, Xinbo Liu, Tongliang Xidian Univ Sch Telecommun Engn State Key Lab Integrated Serv Networks Xian 710071 Peoples R China Xidian Univ Sch Cyber Engn State Key Lab Integrated Serv Networks Xian 710071 Peoples R China Chongqing Univ Posts & Telecommun Chongqing Key Lab Image Cognit Chongqing 400065 Peoples R China Univ Sydney Fac Engn Sydney AI Ctr Sch Comp Sci Darlington NSW 2008 Australia

This paper targets a novel trade-off problem in generalizable prompt learning for vision-language models (VLM), i.e., improving the performance on unseen classes while maintaining the performance on seen classes. Comparing with existing generalizable methods that neglect the seen classes degradation, the setting of this problem is stricter and fits more closely with practical applications. To solve this problem, we start from the optimization perspective, and leverage the relationship between loss landscape geometry and model generalization ability. By analyzing the loss landscapes of the state-of-the-art method and vanilla Sharpness-aware Minimization (SAM) based method, we conclude that the trade-off performance correlates to both loss value and loss sharpness, while each of them is indispensable. However, we find the optimizing gradient of existing methods cannot maintain high relevance to both loss value and loss sharpness during optimization, which severely affects their trade-off performance. To this end, we propose a novel SAM-based method for prompt learning, denoted as Gradient Constrained Sharpness-aware Context Optimization (GCSCoOp), to dynamically constrain the optimizing gradient, thus achieving above two-fold optimization objective simultaneously. Extensive experiments verify the effectiveness of GCSCoOp in the trade-off problem.

关键词： Optimization Minimization Visualization Training Electronic mail Degradation Vectors Telecommunications Intserv networks Geometry Parameter-efficient fine-tuning domain generalization prompt learning vision-language model sharpness-aware minimization

来源：评论

学校读者我要写书评

暂无评论

Influence of structured surface on the interlaminar delamination properties of bonded fiber-metal laminates

引用

THEORETICAL AND APPLIED FRACTURE MECHANICS 2025年 138卷

作者： Yang, Xingyu Liu, Hao Gong, Yu Tian, Dingli Zhang, Jianyu Hu, Ning Chongqing Univ Coll Aerosp Engn Chongqing Key Lab Heterogeneous Mat Mech Chongqing 400044 Peoples R China Chongqing Univ State Key Lab Coal Mine Disaster Dynam & Control Chongqing 400044 Peoples R China Hebei Univ Technol Sch Mech Engn State Key Lab Reliabil & Intelligence Elect Equipm Tianjin 300401 Peoples R China Tongji Univ Civil Aviat Adm China Key Lab Ai aided Airworthiness Civil Aircraft Stru Shanghai Peoples R China

Fiber-metal laminates (FMLs) often exhibit weak interlayer mechanical properties. Delamination is a common issue across FMLs and significantly affects the overall mechanical performance, highlighting the critical need to enhance the interfacial bonding between metal and resin. Structured surface treatment has emerged as a promising method to improve interfacial bonding and has garnered significant attention in adhesive bonding applications. This study investigates the impact of surface structuring on delamination properties by designing two groove patterns (circular and hexagonal) on the metal surface of carbon fiber-reinforced aluminum alloy laminates (CARALL). Specimens were prepared and subsequently subjected to mode I and mode II static delamination tests. Test results indicate that both circular and hexagonal grooves substantially enhance interlaminar fracture toughness, with hexagonal grooves showing a superior improvement. Additionally, finite element models for double cantilever beam (DCB) and end-notched flexure (ENF) tests were developed using the cohesive zone model and progressive damage method. The numerical load-displacement curves were consistent with experimental curves, confirming the accuracy and applicability of the finite element approach.

关键词： Fiber-metal laminates Structured surfaces Delamination growth Fracture toughness Finite element analysis

来源：评论

学校读者我要写书评

暂无评论

Graphics processing unit/artificial neural network-accelerated large-eddy simulation of swirling premixed flames

引用

PHYSICS OF FLUIDS 2024年第5期36卷 055147-055147页

作者： Zhang, Min Mao, Runze Li, Han An, Zhenhua Chen, Zhi X. Peking Univ Coll Engn State Key Lab Turbulence & Complex Syst Beijing 100871 Peoples R China AI Sci Inst AISI Beijing 100080 Peoples R China Xi An Jiao Tong Univ State Key Lab Multiphase Flow Power Engn Xian 710049 Peoples R China

Within the scope of reacting flow simulations, the real-time direct integration (DI) of stiff ordinary differential equations for the computation of chemical kinetics stands as the primary demand on computational resources. Meanwhile, as the number of transport equations that need to be solved increases, the computational cost grows more substantially, particularly for those combustion models involving direct coupling of chemistry and flow such as the transported probability density function model. In the current study, an integrated graphics processing unit-artificial neural network (GPU-ANN) framework is introduced to comply with heavy computational costs while maintaining high fidelity. Within this framework, a GPU-based solver is employed to solve partial differential equations and compute thermal and transport properties, and an ANN is utilized to replace the calculation of reaction rates. Large eddy simulations of two swirling flames provide a robust validation, affirming and extending the GPU-ANN approach's applicability to challenging scenarios. The simulation results demonstrate a strong correlation in the macro flame structure and statistical characteristics between the GPU-ANN approach and the traditional central processing unit (CPU)-based solver with DI. This comparison indicates that the GPU-ANN approach is capable of attaining the same degree of precision as the conventional CPU-DI solver, even in more complex scenarios. In addition, the overall speed-up factor for the GPU-ANN approach is over two orders of magnitude. This study establishes the potential groundwork for widespread application of the proposed GPU-ANN approach in combustion simulations, addressing various and complex scenarios based on detailed chemistry, while significantly reducing computational costs.

关键词： Graphics processing unit

来源：评论

学校读者我要写书评

暂无评论

Learning With Noisy labels Over Imbalanced Subpopulations

引用

IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS 2025年第4期36卷 6544-6555页

作者： Chen, Mingcai Zhao, Yu He, Bing Han, Zongbo Huang, Junzhou Wu, Bingzhe Yao, Jianhua Nanjing Univ State Key Lab Novel Software Technol Nanjing 210000 Peoples R China Tencent AI Lab Shenzhen 518000 Peoples R China Univ Texas Arlington Comp Sci & Engn Dept Arlington TX 76019 USA

Learning with noisy labels (LNL) has attracted significant attention from the research community. Many recent LNL methods rely on the assumption that clean samples tend to have a "small loss." However, this assumption often fails to generalize to some real-world cases with imbalanced subpopulations, that is, training subpopulations that vary in sample size or recognition difficulty. Therefore, recent LNL methods face the risk of misclassifying those "informative" samples (e.g., hard samples or samples in the tail subpopulations) into noisy samples, leading to poor generalization performance. To address this issue, we propose a novel LNL method to deal with noisy labels and imbalanced subpopulations simultaneously. It first leverages sample correlation to estimate samples' clean probabilities for label correction and then utilizes corrected labels for distributionally robust optimization (DRO) to further improve the robustness. Specifically, in contrast to previous works using classification loss as the selection criterion, we introduce a feature-based metric that takes the sample correlation into account for estimating samples' clean probabilities. Then, we refurbish the noisy labels using the estimated clean probabilities and the pseudo-labels from the model's predictions. With refurbished labels, we use DRO to train the model to be robust to subpopulation imbalance. Extensive experiments on a wide range of benchmarks demonstrate that our technique can consistently improve state-of-the-art (SOTA) robust learning paradigms against noisy labels, especially when encountering imbalanced subpopulations. We provide our code in https://***/chenmc1996/LNL-IS.

关键词： Noise measurement Noise Training Tail Optimization Predictive models Estimation Distributionally robust optimization (DRO) label refurbishment learning with noisy labels (LNLs) subpopulation imbalance

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：