检索结果-内蒙古大学图书馆

Skeleton-prompt: A cross-dataset transfer learning approach for skeleton action recognition

Pattern Recognition 2026年 169卷

作者： Mingqi Lu Xiaobo Lu Jun Liu School of Automation Southeast University Nanjing 210096 China Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Nanjing 210096 China School of Computing and Communications Lancaster University UK

This paper presents Skeleton-Prompt, a novel tuning method designed to tackle cross-dataset transfer issues in skeleton action recognition models. Given the scarcity of large-scale 3D skeleton datasets and the variability in keypoint structures across datasets, existing methods often rely on training models from scratch, necessitating extensive labeled data and exhibiting high sensitivity to occlusion. Our approach aims to fine-tune pre-trained models to adapt to limited real-world skeleton data. We use 2D skeletons as inputs and leverage a large human motion dataset for 2D to 3D pose estimation to learn generalizable motion features. A lightweight prompt generator produces instance-level prompts, and we employ dynamic queries with cross-attention to refine the semantic information of the input data. Additionally, we introduce a joint-enhanced multi-stream fusion mechanism based on self-attention to improve robustness against incomplete skeletons. Skeleton-Prompt represents a significant advancement in efficient fine-tuning for skeleton action recognition, effectively addressing cross-dataset generalization challenges in a data-efficient and parameter-efficient manner.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Stochastic DoS attack allocation against collaborative estimation in sensor networks

Stochastic DoS attack allocation against collaborative estim...

引用

作者： Zhang, Ya Du, Lishuang Lewis, Frank L. School of Automation Southeast University Nanjing210096 China Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Nanjing210096 China University of Texas at Arlington Research Institute Fort WorthTX76118 United States

In this paper, denial of service DoS attack management for destroying the collaborative estimation in sensor networks and minimizing attack energy from the attacker perspective is studied. In the communication channels between sensors and a remote estimator, the attacker chooses some channels to randomly jam DoS attacks to make their packets randomly dropped. A stochastic power allocation approach composed of three steps is proposed. Firstly, the minimum number of channels and the channel set to be attacked are given. Secondly, a necessary condition and a sufficient condition on the packet loss probabilities of the channels in the attack set are provided for general and special systems, respectively. Finally, by converting the original coupling nonlinear programming problem to a linear programming problem, a method of searching attack probabilities and power to minimize the attack energy is proposed. The effectiveness of the proposed scheme is verified by simulation examples. © 2014 Chinese Association of Automation.

关键词： Denial-of-service attack

来源：评论

学校读者我要写书评

暂无评论

A NEW IMAGE DENOISING METHOD BASED ON STEER KETNEL REGRESSION

A NEW IMAGE DENOISING METHOD BASED ON STEER KETNEL REGRESSIO...

引用

Jiangsu Annual Conference on Automation (JACA 2020)

作者： G. Shi L. Chai L. Jin Key Laboratory of Measurement and Control of Complex Systems of Engineering and School of Automation Southeast University Nanjing 210096 China

In this paper, we propose a new image pyramid denoising model based on the iterative steer kernel regression (ISKR) method. By introducing an image pyramid to generate images at multiple scales, it can filter out large-scale noise on low-resolution images and preserve details on the high-resolution image, and then we fuse the denoised images of multiple resolutions into a high-quality image by up-sampling noise residual. This method can improve denoising performance efficiently. and reduce the cost of calculation. Experiments show that this model reduces calculation by about 58% compared with the ISKR method while ensuring a good denoising effect.

关键词： image resolution image denoising regression analysis

来源：评论

学校读者我要写书评

暂无评论

Intermittent Formation control of Multi-agent systems under Strongly Connected Communication Topologies

Intermittent Formation Control of Multi-agent Systems under ...

引用

International Conference on control and Automation (ICCA)

作者： Ke Zhou Yang-Yang Chen Faxiang Zhang School of Automation and Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Southeast University Nanjing China

ISBN: (数字)9781728190938

ISBN: (纸本)9781728190945

This paper addresses the intermittent formation control problem of multi-agent systems (MASs) under directed topologies. To achieve the ideal formation effectively, (i.e., tracking a given set of closed orbits) when the communication is intermittent. The intermittent control approach is applied to deal with such situation. The main contribution is that the proposed controller can achieve global stable formation in the Lyapunov sense based on the consensus strategy. The simulation results obtained for an example demonstrate the feasibility and effectiveness of the proposed method (i.e, if the time length of communication is larger than a threshold value, the formation can be realized).

关键词： Orbits Mathematical model Topology Target tracking Stability analysis Trajectory Switches

来源：评论

学校读者我要写书评

暂无评论

Optimization Method of Target Detection and Tracking System for Mobile Robot

Optimization Method of Target Detection and Tracking System ...

引用

IEEE International Conference on Real-time Computing and Robotics (RCAR)

作者： Ziqi Liu Fang Fang Kun Qian Bo Zhou Yahui Gan Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education School of Automation Southeast University Nanjing China

ISBN: (数字)9781728172934

ISBN: (纸本)9781728172941

The existing target detection and tracking methods, as an underlying design of service robot, are highly dependent on the computing ability, which will limit the transplantation to the robot platform. To solve the problem, this paper proposes a detection and tracking system suitable for mobile robot. For the target detection, an optimization method for feature extraction is proposed, by reducing the dimensionality of feature vectors to improve detection speed. For the target tracking, a multi-scale optimization method is proposed, which can accurately obtain all the characteristics of the upper body of the target. The last section is the experiment of the integrated system, which will resist the interference and robustly dynamically tracking the target.

关键词： Target tracking Feature extraction Robots Optimization methods Real-time systems Object detection Histograms

来源：评论

学校读者我要写书评

暂无评论

Spherical Formation Tracking control of Spacecraft Flying Under Directed Communication Topology and External Disturbance

Spherical Formation Tracking Control of Spacecraft Flying Un...

引用

International Conference on control and Automation (ICCA)

作者： Rui Yu Yang-Yang Chen Fa-Xiang Zhang School of Automation and Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Southeast University Nanjing China

ISBN: (数字)9781728190938

ISBN: (纸本)9781728190945

This paper addresses the formation tracking control problem for spacecrafts flying spherically with certain formation. Compared to existing works that deal with bidirectional communication topology and disturbance-free scenarios, this paper proposes a brand-new formation tracking control protocol with directed network and unknown external disturbance with a known upperbound. Moreover, spacecraft flying model is introduced to the system dynamics. The formation tracking law of spacecraft is designed, and the uniform ultimate boundedness is proven through Lyapunov functions as well. To avoid the "explosion of terms" caused by time derivative, the dynamic surface control (DSC) technology is employed to the designing procedure of control input. Numerical simulation results are presented to demonstrate the effectiveness of the proposed scheme.

关键词： Space vehicles Orbits Topology Trajectory Explosions Aerospace electronics Vehicle dynamics

来源：评论

学校读者我要写书评

暂无评论

An Event-Triggered χ2-Detector for Cyber-Physical systems under False Data Injection Attacks

An Event-Triggered χ2-Detector for Cyber-Physical Systems u...

引用

International Conference on control and Automation (ICCA)

作者： Guoxi Chen Ya Zhang School of Automation Southeast University Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Nanjing China

ISBN: (数字)9781728190938

ISBN: (纸本)9781728190945

Attacks on the cyber-physical systems (CPSs) are increasing recently, which makes the research of CPSs' security issues more urgent. χ 2 -detector, the most common attack detector, has a fatal defect, that it requires all the covariance matrices of observation residual be reversible. In this paper, we propose a novel method to construct an attack detection variable obeying χ 2 distribution without special requirements, and develop an event-triggered mechanism to reduce the adverse impact of the detection error. Observer-based state feedback controller is applied and a sufficient condition of stability on the alarm probability of the detector and the parameter in the event-triggered mechanism are further given. Finally, a simulation example is given to verify the effectiveness of the proposed attack detector.

关键词： Detectors Covariance matrices Stability analysis control systems Time measurement Security Mathematical model

来源：评论

学校读者我要写书评

暂无评论

Consensus Kalman Filtering for Sensor Networks Based on FDI Attack Detection

Consensus Kalman Filtering for Sensor Networks Based on FDI ...

引用

International Conference on control, Automation, Robotics and Vision (ICARCV)

作者： Jiali Hao Ya Zhang School of Automation Southeast University and with Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Nanjing China

ISBN: (数字)9781728177090

ISBN: (纸本)9781728177106

This paper studies the consensus Kalman filtering algorithm with distributed attack detection for reducing the effects of false data injection attacks in wireless sensor networks. The FDI attacks are randomly injected into communication channels or sensors with certain probabilities, which undermines the accuracy of the transmission data and the accuracy of measurement data respectively. χ 2 detector is applied, and if the received data is determined to be attacked, it is omitted in consensus Kalman filtering. It is proved that if the FDI attack is randomly injected into communication channels, under the consensus Kalman filtering algorithm with adaptive weighting protocol and attack detection, the estimation errors of the sensor network can be bounded in probability. If the FDI attack is randomly injected into sensors, a probability condition on the attacks is given to guarantee the estimation errors of the sensor network bounded in mean square sense. Numerical simulations are conducted to demonstrate the performance of the proposed algorithms.

关键词： Covariance matrices Kalman filters State estimation Wireless sensor networks Robot sensing systems Distributed databases Detectors

来源：评论

学校读者我要写书评

暂无评论

Improved Synthesis of Nondeterministic Publicly-Known Edit Functions for Opacity Enforcement

引用

IEEE Transactions on Automatic control 2025年

作者： Liu, Rongjian Hadjicostis, Christoforos N. Chongqing Normal University National Center for Applied Mathematics in Chongqing Chongqing401331 China Southeast University Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Nanjing210096 China University of Cyprus Department of Electrical and Computer Engineering Nicosia Cyprus

In this paper, we investigate the synthesis problem of edit functions for opacity enforcement in systems modelled as partially-observed finite-state automata. For better plausible deniability for the edit functions, in cases where the intruder may learn/know their existence, a synthesis problem for nondeterministic edit functions under a publicly-known setting was introduced in the literature, but not yet fully solved. In light of this, we revisit the publicly-known edit mechanism by proposing a sound and complete approach for solving the synthesis problem of nondeterministic publicly-known edit functions. For practical applications, the edit function is assumed to have limited erasure and insertion ability. Our approach is based on a new labelled finite-state structure, and results in wider applicability than previously-proposed in the literature. © 1963-2012 IEEE.

关键词： Discrete event simulation

来源：评论

学校读者我要写书评

暂无评论

Knowledge Induced Deep Q-Network for Robot Push and Grasp Manipulation Skills Learning

Knowledge Induced Deep Q-Network for Robot Push and Grasp Ma...

引用

Chinese Automation Congress (CAC)

作者： Boxing Gui Kun Qian Shenghao Chen Xingshuo Jing Xudong Ma School of Automation Southeast University Nanjing China Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education of China

ISBN: (数字)9781728176871

ISBN: (纸本)9781728176888

In the process of robotic push and grasp manipulation skills learning, the prior knowledge of object shape and environmental constraints can be employed to make corresponding manipulation policies. In this paper, a DQN decision model is proposed for learning multi-step push and grasp manipulation of different types of objects. The key of the proposed method is that the prior knowledge of object shape and environmental constraints is introduced into the reward system by a SVM classification model. Therefore, the reward system will continuously motivate the robot to approach to the target manipulation during training process. Simulated and actual verification of the trained model is carried out. The experimental results show that the trained model has a high success rate in grasping cube objects, and it can also successfully complete the task of pushing cylindrical objects to the constraint boundary.

关键词： Robots Support vector machines Training Shape Robot sensing systems Reinforcement learning Grasping

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：