Journal of The Franklin Institute-engineering and Applied Mathematics最新文献_第8页

Face recognition method based on fusion of improved MobileFaceNet and adaptive Gamma algorithm 基于改进型 MobileFaceNet 和自适应伽马算法融合的人脸识别方法

IF 3.7 3区计算机科学

Journal of The Franklin Institute-engineering and Applied Mathematics Pub Date : 2024-09-27 DOI: 10.1016/j.jfranklin.2024.107306

Jingwei Li, Yipei Ding, Zhiyu Shao, Wei Jiang

{"title":"Face recognition method based on fusion of improved MobileFaceNet and adaptive Gamma algorithm","authors":"Jingwei Li, Yipei Ding, Zhiyu Shao, Wei Jiang","doi":"10.1016/j.jfranklin.2024.107306","DOIUrl":"10.1016/j.jfranklin.2024.107306","url":null,"abstract":"<div><div>MobileFaceNet face recognition algorithm is a relatively mainstream face recognition algorithm at present. Its advantages of small memory and fast running speed make it widely used in embedded devices. Due to the limited face image acquisition capability of embedded devices, the accuracy of face recognition is often reduced due to uneven illumination and poor exposure quality. In order to solve this problem, a face recognition algorithm based on the fusion of MobileFaceNet and adaptive Gamma algorithm is proposed. The application of the algorithm proposed in this paper in image preprocessing is as follows. Firstly, adaptive Gamma correction is used to improve the brightness of the face image. Then, the edge of the face image is enhanced by the Laplace operator. Finally, a linear weighted fusion was performed between the Gamma corrected image and the enhanced edge image to obtain the pre-processed face image. At the same time, we have improved the traditional MobileFaceNet network. The feature extraction network MobileFaceNet has been improved by adding a Stylebased Recall Module (SRM) attention mechanism to its bottom neck layer, utilizing the mean and standard deviation of input features to improve the ability to capture global information and enhance more important feature information. Finally, the proposed method was verified on the LFW and Agedb face test set. The experimental results show that the adaptive Gamma algorithm proposed in this paper and the improvement of MobileFaceNet can achieve a face recognition accuracy of 99.27 % on LFW dataset and 90.18 % on Agedb dataset while only increasing the model size by 0.4 M and the processing speed for each image is enhanced by 4 ms. which can effectively improve the accuracy of face recognition and better application prospects on embedded devices. The method presented in this article has certain practical significance.</div></div>","PeriodicalId":17283,"journal":{"name":"Journal of The Franklin Institute-engineering and Applied Mathematics","volume":"361 17","pages":"Article 107306"},"PeriodicalIF":3.7,"publicationDate":"2024-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142420817","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Two improved generalized extended stochastic gradient algorithms for CARARMA systems CARARMA 系统的两种改进型广义扩展随机梯度算法

IF 3.7 3区计算机科学

Journal of The Franklin Institute-engineering and Applied Mathematics Pub Date : 2024-09-26 DOI: 10.1016/j.jfranklin.2024.107295

Lingling Lv , Yulin Zhang , Quanzhen Huang , Yu Wu

引用次数: 0

Spatial–Temporal Similarity Fusion Graph Adversarial Convolutional Networks for traffic flow forecasting 用于交通流量预测的时空相似性融合图对抗卷积网络

IF 3.7 3区计算机科学

Journal of The Franklin Institute-engineering and Applied Mathematics Pub Date : 2024-09-26 DOI: 10.1016/j.jfranklin.2024.107299

Bin Wang, Zhendan Long, Jinfang Sheng, Qiang Zhong

{"title":"Spatial–Temporal Similarity Fusion Graph Adversarial Convolutional Networks for traffic flow forecasting","authors":"Bin Wang, Zhendan Long, Jinfang Sheng, Qiang Zhong","doi":"10.1016/j.jfranklin.2024.107299","DOIUrl":"10.1016/j.jfranklin.2024.107299","url":null,"abstract":"<div><div>Traffic flow forecasting is integral to the advancement of intelligent transportation systems and the development of smart cities. This paper introduces a novel model, the Spatial–Temporal Similarity Fusion Graphs Adversarial Convolutional Networks (STSF-GACN), which leverages advanced data preprocessing techniques to enhance the predictive accuracy and efficiency of traffic flow forecasting.</div><div>The innovation of our approach lies in the meticulous construction of the spatial–temporal similarity matrix through the precise calculation of temporal and spatial similarities. This matrix forms the backbone of our model, serving as the generator in the integrated Generative Adversarial Network (GAN) architecture. The Spatial–Temporal Similarity Fusion Adaptive Graph Convolutional Network, developed as part of our GAN’s generator, utilizes cutting-edge techniques such as the Wasserstein distance and Dynamic Time Warping to optimize the adaptive adjacency matrix, enabling the model to capture latent spatial–temporal correlations with unprecedented depth and precision.</div><div>The discriminator of the GAN further refines the model by evaluating the accuracy of the traffic predictions, ensuring that the generative model produces results that are not only accurate but also robust against varying traffic conditions. This cohesive integration of GAN into the model architecture allows for a significant improvement in prediction accuracy and convergence speed, moving beyond traditional forecasting methods.</div></div>","PeriodicalId":17283,"journal":{"name":"Journal of The Franklin Institute-engineering and Applied Mathematics","volume":"361 17","pages":"Article 107299"},"PeriodicalIF":3.7,"publicationDate":"2024-09-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142420682","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A Lyapunov–Razumikhin control strategy for stochastic nonlinear delayed systems with polynomial conditions 具有多项式条件的随机非线性延迟系统的 Lyapunov-Razumikhin 控制策略

IF 3.7 3区计算机科学

Journal of The Franklin Institute-engineering and Applied Mathematics Pub Date : 2024-09-26 DOI: 10.1016/j.jfranklin.2024.107288

Han-Wen Zhang , Zhen-Guo Liu , Wei-Hai Zhang

引用次数: 0

Reachable set control of singular semi-Markov jump multi-agent systems with input delay via state decomposition method 通过状态分解法实现具有输入延迟的奇异半马尔可夫跃迁多代理系统的可达集控制

IF 3.7 3区计算机科学

Journal of The Franklin Institute-engineering and Applied Mathematics Pub Date : 2024-09-25 DOI: 10.1016/j.jfranklin.2024.107294

Mingde Liu , Liang Zhang , Ning Zhao , Sanaa Sharaf , Lun Li

引用次数: 0

Event-triggered finite-time fault-tolerance control and simultaneous disturbance rejection for Markov jump systems with general transition probabilities 具有一般过渡概率的马尔可夫跃迁系统的事件触发有限时间容错控制和同步干扰抑制

IF 3.7 3区计算机科学

Journal of The Franklin Institute-engineering and Applied Mathematics Pub Date : 2024-09-24 DOI: 10.1016/j.jfranklin.2024.107292

Xiang Pan , Guochen Pang , Xiangyong Chen , Yang Liu , Jianlong Qiu , Jinde Cao

引用次数: 0

Output synchronization of a class of complex dynamic networks: A reinforcement learning method 一类复杂动态网络的输出同步：一种强化学习方法

IF 3.7 3区计算机科学

Journal of The Franklin Institute-engineering and Applied Mathematics Pub Date : 2024-09-24 DOI: 10.1016/j.jfranklin.2024.107284

Ning Zheng, Jinxu Liu, Lei Su, Shaoyu Lv, Hao Shen

引用次数: 0

Fusion estimation for state-saturated multi-rate systems with integral measurement and sensor resolution 具有积分测量和传感器分辨率的状态饱和多速率系统的融合估计

IF 3.7 3区计算机科学

Journal of The Franklin Institute-engineering and Applied Mathematics Pub Date : 2024-09-24 DOI: 10.1016/j.jfranklin.2024.107281

Long Xu , Changcheng Shi , Hui Yu , Ling Hou

{"title":"Fusion estimation for state-saturated multi-rate systems with integral measurement and sensor resolution","authors":"Long Xu , Changcheng Shi , Hui Yu , Ling Hou","doi":"10.1016/j.jfranklin.2024.107281","DOIUrl":"10.1016/j.jfranklin.2024.107281","url":null,"abstract":"<div><div>In this paper, the fusion estimation problem is studied for a class of multi-rate systems (MRS) subject to state saturation. The state update rate of the system and the sampling rate of sensors are different in MRS. In addition, to incorporate the requirements of actual engineering practice, the sensor resolution and the phenomenon of integral measurements are considered in the measurement output of sensors. The aim of this paper is to design an appropriate set of local estimators for each sensor node by taking into account the effects of integral measurement, sensor resolution and multi-rate sampling mechanism. Then, the upper bound on the error covariance of the estimator can be obtained by calculating solution of the matrix difference equation and the relevant filter parameter is determined by minimizing the upper bound. Based on the covariance intersection (CI) method, the fusion estimation is obtained by fusing local estimates. The consistency is demonstrated for this fusion estimating approach. Finally, the effectiveness and appropriateness of the multi-rate fusion estimation approach are confirmed by comprehensive simulations.</div></div>","PeriodicalId":17283,"journal":{"name":"Journal of The Franklin Institute-engineering and Applied Mathematics","volume":"361 18","pages":"Article 107281"},"PeriodicalIF":3.7,"publicationDate":"2024-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142420763","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Projective lag synchronization of fractional delayed memristive neural networks via event-based hybrid pinning impulsive controller 通过基于事件的混合针刺脉冲控制器实现分数延迟记忆神经网络的投影滞后同步

IF 3.7 3区计算机科学

Journal of The Franklin Institute-engineering and Applied Mathematics Pub Date : 2024-09-24 DOI: 10.1016/j.jfranklin.2024.107297

Huiyu Wang , Shutang Liu , Xiang Wu , Wei Qiao , Jie Sun

{"title":"Projective lag synchronization of fractional delayed memristive neural networks via event-based hybrid pinning impulsive controller","authors":"Huiyu Wang , Shutang Liu , Xiang Wu , Wei Qiao , Jie Sun","doi":"10.1016/j.jfranklin.2024.107297","DOIUrl":"10.1016/j.jfranklin.2024.107297","url":null,"abstract":"<div><div>This paper delves into the projective lag synchronization of Riemann–Liouville type fractional-order memristive neural networks accounting for jump mismatch. Recognizing the inherent inconsistencies in synchronizing traditional fractional-order memristive neural networks, we introduce a novel mathematical model that accommodates the jump mismatch phenomenon. A groundbreaking event-based hybrid pinning impulsive controller is then introduced, equipped with tailored event-triggering conditions, to elucidate the global asymptotic projective lag synchronization. Leveraging inequality principles and impulse analysis, a new Lyapunov functional is proposed, formulating sufficient conditions for synchronization while theoretically eliminating Zeno behavior in the controller. Notably, our approach substantially optimizes control overhead while fulfilling practical synchronization requisites. In addition, the obtained sufficient conditions can theoretically guide practical engineering applications of the network. Finally, a simulation example, emphasizing varied projective and lag factors, demonstrates our findings.</div></div>","PeriodicalId":17283,"journal":{"name":"Journal of The Franklin Institute-engineering and Applied Mathematics","volume":"361 18","pages":"Article 107297"},"PeriodicalIF":3.7,"publicationDate":"2024-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142322459","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Prescribed-time practical stabilization via majorant systems 通过主系统实现规定时间内的实际稳定

IF 3.7 3区计算机科学

Journal of The Franklin Institute-engineering and Applied Mathematics Pub Date : 2024-09-23 DOI: 10.1016/j.jfranklin.2024.107298

Laura Celentano , Alison Garza-Alonso , Michael V. Basin

{"title":"Prescribed-time practical stabilization via majorant systems","authors":"Laura Celentano , Alison Garza-Alonso , Michael V. Basin","doi":"10.1016/j.jfranklin.2024.107298","DOIUrl":"10.1016/j.jfranklin.2024.107298","url":null,"abstract":"<div><div>This paper provides a design methodology of simple control laws to solve the prescribed-time practical stabilization problem for a class of pseudo-linear systems with a single input, also unstable, subject to perturbations, nonlinear and discontinuous, but bounded. The perturbations on the dynamic matrix, such as a ratio of multi-affine polynomials of nonlinear functions or parameters, are not triangular, and a perturbation on the input of multiplicative type is also considered. The stated methodology is based on the majorant systems approach with optimized quadratic Lyapunov functions. The main advantages of the proposed control laws consist in the simplicity of their design and implementation and the reliable performance. Indeed, these control laws depend on a single design parameter, have constant gains, and guarantee the practical convergence with a prescribed relative error, also after a prescribed time. Some examples, also an engineering one, are provided to validate the obtained theoretical results and show the effectiveness and efficiency of the proposed control laws.</div></div>","PeriodicalId":17283,"journal":{"name":"Journal of The Franklin Institute-engineering and Applied Mathematics","volume":"361 18","pages":"Article 107298"},"PeriodicalIF":3.7,"publicationDate":"2024-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142663678","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0