检索结果-内蒙古大学图书馆

Geometry-aware 3D pose transfer using transformer autoencoder

Computational Visual Media 2024年第6期10卷 1063-1078页

作者： Shanghuan Liu Shaoyan Gai Feipeng Da Fazal Waris Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of EducationSoutheast UniversityNanjing210096China

3D pose transfer over unorganized point clouds is a challenging generation task,which transfers a source’s pose to a target shape and keeps the target’s *** deep models have learned deformations and used the target’s identity as a style to modulate the combined features of two shapes or the aligned vertices of the source ***,all operations in these models are point-wise and independent and ignore the geometric information on the surface and structure of the input *** disadvantage severely limits the generation and generalization *** this study,we propose a geometry-aware method based on a novel transformer autoencoder to solve this *** efficient self-attention mechanism,that is,cross-covariance attention,was utilized across our framework to perceive the correlations between points at different ***,the transformer encoder extracts the target shape’s local geometry details for identity attributes and the source shape’s global geometry structure for pose *** transformer decoder efficiently learns deformations and recovers identity properties by fusing and decoding the extracted features in a geometry attentional manner,which does not require corresponding information or modulation *** experiments demonstrated that the geometry-aware method achieved state-of-the-art performance in a 3D pose transfer *** implementation code and data are available at https://***/SEULSH/Geometry-Aware-3D-Pose-Transfer-Using-Transfor mer-Autoencoder.

关键词： 3D pose transfer geometry-aware transfor-mer autoencoder cross-covariance attention

来源：评论

学校读者我要写书评

暂无评论

Regional Multi-Agent Cooperative Reinforcement Learning for City-Level Traffic Grid Signal control

引用

IEEE/CAA Journal of Automatica Sinica 2024年第9期11卷 1987-1998页

作者： Yisha Li Ya Zhang Xinde Li Changyin Sun School of Automation Southeast UniversityNanjing 210096 Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of EducationSoutheast UniversityNanjing 210096China

This article studies the effective traffic signal control problem of multiple intersections in a city-level traffic system.A novel regional multi-agent cooperative reinforcement learning algorithm called RegionSTLight is proposed to improve the traffic *** a regional multi-agent Q-learning framework is proposed,which can equivalently decompose the global Q value of the traffic system into the local values of several regions Based on the framework and the idea of human-machine cooperation,a dynamic zoning method is designed to divide the traffic network into several strong-coupled regions according to realtime traffic flow *** order to achieve better cooperation inside each region,a lightweight spatio-temporal fusion feature extraction network is *** experiments in synthetic real-world and city-level scenarios show that the proposed RegionS TLight converges more quickly,is more stable,and obtains better asymptotic performance compared to state-of-theart models.

关键词： Human-machine cooperation mixed domain attention mechanism multi-agent reinforcement learning spatio-temporal feature traffic signal control

来源：评论

学校读者我要写书评

暂无评论

Context-Aided Semantic-Aware Self-Alignment for Video-Based Person Re-Identification

引用

IEEE Transactions on Circuits and systems for Video Technology 2025年第6期35卷 5307-5317页

作者： Ran, Zhidan Xiao, Zhiyao Lu, Xiaobo Wei, Xuan Liu, Wei Southeast University School of Automation Nanjing210096 China Ministry of Education Key Laboratory of Measurement and Control of Complex Systems of Engineering Nanjing210096 China

Video-based person re-identification (Re-ID) aims at associating the video sequences of the identical person across multiple cameras. The ubiquitous appearance misalignment poses a major obstacle for video person Re-ID. Existing alignment-based methods generally rely on off-the-shelf semantic parsing models to locate visible human parts, which ignore identifiable personal belongings and cannot handle various interferences (e.g., pedestrian detection errors and occlusions) in video clips. In this work, we propose a novel framework termed Context-Aided Semantic-Aware Self-Alignment (CSSA) for video-based person Re-ID. First, we propose to jointly learn pixel-level part-aligned representations and semantic-aligned global-level representations in an end-to-end manner. Unlike most existing approaches that depend on prior information in terms of pose for part estimation, CSSA can locate different body parts and achieve the pixel-level semantic alignment without extra human topology semantics. Second, a Context-Aided Region Enhancement (CARE) module is proposed to efficiently highlight macro-visual patterns associated with the target pedestrian and suppress noise caused by factors like background clutters and occlusions. Third, we propose a Semantic-Aware Global Feature Alignment (SGFA) method for generating pair-wise semantic-aligned global representations, which play an essential role in both the training and inference phases. Extensive experimental results on multiple challenging benchmarks indicate the superiority and effectiveness of the proposed CSSA. © 1991-2012 IEEE.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

Disturbance Observer-based Pointing control of Leighton Chajnantor Telescope

引用

Research in Astronomy and Astrophysics 2024年第1期24卷 135-150页

作者： Weirui Chen Zheng Wang School of Automation Southeast UniversityNanjing 210096China Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of EducationSoutheast UniversityNanjing 210096China

Leighton Chajnantor Telescope(LCT), i.e., the former Caltech Submillimeter Observatory telescope, will be refurbished at the new site in Chajnantor Plateau, Chile in 2023. The environment of LCT will change significantly after its relocation, and the telescope will be exposed to large wind disturbances directly because its enclosure will be completely open during observation. The wind disturbance is expected to be a challenge for LCT's pointing control since the existing control method cannot reject this disturbance very well. Therefore, it is very necessary to develop a new pointing control method with good capability of disturbance rejection. In this research, a disturbance observer—based composite position controller(DOB-CPC) is designed, in which an H∞feedback controller is employed to compress the disturbance, and a feedforward linear quadratic regulator is employed to compensate the disturbance precisely based on the estimated disturbance signal. Moreover, a controller switching policy is adopted, which applies the proportional controller to the transient process to achieve a quick response and applies the DOB-CPC to the steady state to achieve a small position error. Numerical experiments are conducted to verify the good performance of the proposed pointing controller(i.e., DOB-CPC) for rejecting the disturbance acting on LCT.

关键词： instrumentation:miscellaneous methods:analytical methods:numerical telescopes

来源：评论

学校读者我要写书评

暂无评论

Balanced prioritized experience replay in off-policy reinforcement learning

引用

Neural Computing and Applications 2024年第25期36卷 15721-15737页

作者： Lou, Zhouwei Wang, Yiye Shan, Shuo Zhang, Kanjian Wei, Haikun Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Nanjing210096 China School of Automation Southeast University Nanjing210096 China

In Off-Policy reinforcement learning (RL), the experience imbalance problem can affect learning performance. The experience imbalance problem refers to the phenomenon that the experiences obtained by the agent during the learning process are unevenly distributed in the state space, resulting in the agent’s inability to accurately estimate the value of each potential state. This problem is typically caused by environments with high-dimensional state and action spaces, as well as the exploration–exploitation mechanism inherent in RL. This article proposes a balanced prioritized experience replay (BPER) algorithm based on experience rarity. First, an evaluation metric to quantify experience rarity is defined. Then, the sampling priority of each experience is calculated according to this metric. Finally, prioritized experience replay is performed according to the sampling priority. BPER increases the sampling frequency of high-rarity experiences and decreases the sampling frequency of low-rarity experiences, enabling the agent to learn more comprehensive knowledge. We evaluate BPER on a series of MuJoCo continuous control tasks. Experimental results show that BPER can effectively improve the learning performance while mitigating the impact of the experience imbalance problem. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

A deep reinforcement learning model for large-scale traffic signal control based on graph meta-learning using local subgraphs

引用

Science China(Information Sciences) 2025年

作者： Zhicheng ZHOU Ya ZHANG Xinde LI School of Automation Southeast University Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education

This paper investigates the problem of traffic signal control in large-scale road networks. A deep reinforcement learning model based on graph meta-learning using local subgraphs is proposed to control the traffic signal. The entire traffic network is represented as a graph by defining traffic lights as nodes and treating connections between intersections as edges. A graph neural network is used to enhance cooperation and communications between agents since information about neighbors is aggregated. To overcome the challenges in large-scale road networks, the proposed model employs a graph neural network on local subgraphs to reduce the difficulty of training in large-scale road networks. The model trained in small-scale traffic networks is transferred to a large-scale traffic network. Agent knowledge acquired from local subgraphs during the training of a small-scale road network confers advantages to the training of large-scale road networks under the resemblance between the structures of local subgraphs in small-and large-scale road networks. Furthermore, meta-learning is used to facilitate the model's rapid adaptability to unseen large-scale road networks. The advantage of the double Q-learning network is taken to reduce overestimation. In experiments, real-world road networks and synthetic road networks comprising more than 1000 intersections are given to evaluate the effectiveness of the proposed model.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Reduction of data amount in data-driven design of linear quadratic regulators

引用

control Theory and Technology 2024年第4期22卷 532-542页

作者： Shinsaku Izumi Xin Xin School of Systems Engineering Kochi University of TechnologyKamiKochi782-8502Japan School of Automation Southeast Universityand Key Laboratory of Measurement and Control of Complex Systems of EngineeringMinistry of EducationNanjing210096JiangsuChina

This paper discusses the data-driven design of linear quadratic regulators,i.e.,to obtain the regulators directly from experimental data without using the models of *** particular,we aim to improve an existing design method by reducing the amount of the required experimental *** the data amount leads to the cost reduction of experiments and computation for the data-driven *** present a simplified version of the existing method,where parameters yielding the gain of the regulator are estimated from only part of the data required in the existing *** then show that the data amount required in the presented method is less than half of that in the existing method under certain *** addition,assuming the presence of measurement noise,we analyze the relations between the expectations and variances of the estimated parameters and the *** a result,it is shown that using a larger amount of the experimental data might mitigate the effects of the noise on the estimated *** results are verified by numerical examples.

关键词： Data-driven design Linear quadratic regulators Linear systems Riccati equation Stochastic properties

来源：评论

学校读者我要写书评

暂无评论

Analysis of the optimal target node to reduce seizure-like discharge in networks

引用

Chinese Physics B 2024年第5期33卷 712-720页

作者：闫璐瑶张红慧孙中奎 School of Mathematics and Statistics Northwestern Polytechnical UniversityXi'an 710129China MIIT Key Laboratory of Dynamics and Control of Complex Systems Xi'an 710129China

Network approaches have been widely accepted to guide surgical strategy and predict outcome for epilepsy *** study starts with a single oscillator to explore brain activity,using a phenomenological model capable of describing healthy and epileptic *** ictal number of seizures decreases or remains unchanged with increasing the speed of oscillator excitability and in each seizure,there is an increasing tendency for ictal duration with respect to the *** underlying reason is that the strong excitability speed is conducive to reduce transition behaviors between two attractor ***,the selection of the optimal removal node is estimated by an indicator proposed in this *** show that when the indicator is less than the threshold,removing the driving node is more possible to reduce seizures significantly,while the indicator exceeds the threshold,the epileptic node could be the removal ***,the driving node is such a potential target that stimulating it is obviously effective in suppressing seizure-like activity compared to other nodes,and the propensity of seizures can be reduced 60%with the increased stimulus *** results could provide new therapeutic ideas for epilepsy surgery and neuromodulation.

关键词： epilepsy driving node epileptic node seizure regulation

来源：评论

学校读者我要写书评

暂无评论

Aerial target threat assessment based on gated recurrent unit and self-attention mechanism

引用

Journal of systems Engineering and Electronics 2024年第2期35卷 361-373页

作者： CHEN Chen QUAN Wei SHAO Zhuang School of Automation Beijing Institute of TechnologyBeijing 100081China State Key Laboratory of Intelligent Control and Decision of Complex Systems Beijing 100081China

Aerial threat assessment is a crucial link in modern air combat, whose result counts a great deal for commanders to make decisions. With the consideration that the existing threat assessment methods have difficulties in dealing with high dimensional time series target data, a threat assessment method based on self-attention mechanism and gated recurrent unit(SAGRU) is proposed. Firstly, a threat feature system including air combat situations and capability features is established. Moreover, a data augmentation process based on fractional Fourier transform(FRFT) is applied to extract more valuable information from time series situation features. Furthermore, aiming to capture key characteristics of battlefield evolution, a bidirectional GRU and SA mechanisms are designed for enhanced ***, after the concatenation of the processed air combat situation and capability features, the target threat level will be predicted by fully connected neural layers and the softmax classifier. Finally, in order to validate this model, an air combat dataset generated by a combat simulation system is introduced for model training and testing. The comparison experiments show the proposed model has structural rationality and can perform threat assessment faster and more accurately than the other existing models based on deep learning.

关键词： target threat assessment gated recurrent unit(GRU) self-attention(SA) fractional Fourier transform(FRFT)

来源：评论

学校读者我要写书评

暂无评论

RSD-Diff: Boundary-Frequency Feature-Enhanced Rail Surface Defect Inspection with Diffusion-Based Transformer Decoder 10

RSD-Diff: Boundary-Frequency Feature-Enhanced Rail Surface D...

引用

10th International Conference on systems and Informatics, ICSAI 2024

作者： Wang, Dezhou Su, Shixiang Lu, Xiaobo School of Automation Southeast University Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Nanjing China

ISBN: (纸本)9798331510138

Rail surface defect inspection is an essential task for the railway system. However, due to the similarity of the background and defect foreground pixels, uneven textures, irregular shapes and multiple scales of the existing defects, the inspection accuracy is still required to be further improved. In this paper, we first expand our dataset by GAN-based method for the scarcity of the defect samples. Then, we propose a diffusion-based model to perform the classification and segmentation task. The diffusion model predicts the mask from the Gaussian noise broken ground truth by successive designed transformer decoder. Meanwhile, to better distinguish the pixels in the boundary or other illegible regions, a boundary-frequency feature enhancement module (BFM) is proposed to make the model pay more attention to the defect-related features. Experiments on the rail surface defect dataset revealed that the proposed diffusion-based transformer decoder and BFM can work well together and efficiently improve the rail surface defect segmentation accuracy. © 2024 IEEE.

关键词： Decoding

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：