Thanks to the biomimetic properties of synaptic plasticity, memristors are often utilized to mimic biological neuronal synapses. This article presents a new memristor synapse coupling (MSC) approach for producing mult...
详细信息
Thanks to the biomimetic properties of synaptic plasticity, memristors are often utilized to mimic biological neuronal synapses. This article presents a new memristor synapse coupling (MSC) approach for producing multidirectional multidouble-scroll attractors. Through adopting flux-controlled hyperbolic memristor synapses to couple a Hopfield neural network, a novel multidirectional multidouble-scroll Hopfield neural network (MDMDSHNN) is constructed. Theoretical results and numerical calculations indicate that MDMDSHNN is capable of producing any desired amount of multidirectional multidouble-scroll attractors, including unidirectional (1-D), bidirectional (2-D), and three-directional (3-D) multidouble-scroll attractors. Furthermore, an infinite amount of initial offset-boosted coexisting multidouble-scroll chaotic attractors possessing identical shapes but different positions, i.e., homogeneous extreme multistability are also found via switching the memristor initial values. Furthermore, to validate the physical implementability and practicality of MDMDSHNN, the digital hardware platform is performed. Finally, to investigate MDMDSHNN in practical application, an image encryption scheme with superior security performance is given by employing the homogeneous multidouble-scroll chaotic sequences, further illustrating good superiority and effectiveness of the present MSC method.
In this paper, we revisit the drive-response synchronization of a class of recurrent neural networks with unbounded delays and time-varying coefficients, contrary to usual in the literature about time-varying neural n...
详细信息
In this paper, we revisit the drive-response synchronization of a class of recurrent neural networks with unbounded delays and time-varying coefficients, contrary to usual in the literature about time-varying neural networks, the signs of self-feedback coefficients are permitted to be indefinite or the time varying coefficients can be unbounded. A generalized scalar delay differential inequality considering indefinite self-feedback coefficient and unbounded delay simultaneously is established, which covers the existing result with bounded delay, the applicabilities of the sufficient conditions are discussed. Some novel criteria for network synchronization are then derived by constructing different candidate functions. These results have been improved in some aspects compared with the existing ones. Differential inequality in vector form is also derived to obtain a more refined synchronization criterion which removes some strong assumptions. Three examples are presented to verify the effectiveness and show the superiorities of our theoretical results. (C) 2021 Elsevier Ltd. All rights reserved.
In order to construct the high-dimensional discrete hyperchaotic systems systemically, this paper proposes a new coupled chaotic model. It has a wide chaotic parameter range and can relax the election of the coupling ...
详细信息
In order to construct the high-dimensional discrete hyperchaotic systems systemically, this paper proposes a new coupled chaotic model. It has a wide chaotic parameter range and can relax the election of the coupling coefficient. Sufficient conditions are derived to prove the existence of Li-Yorke chaos in the proposed model. Meanwhile, the existence of hyperchaos is also demonstrated. To discern the effects of different coupling types on the chaotic dynamics more comprehensively, we further explore the dynamical behaviors with various coupling structures by using Lyapunov spectrum, bifurcation analysis, and phase portraits. We investigate the interaction relationship between coupled units and give suggestions for selecting coupling types. The results indicate that the coupled model has more complex and more stable chaotic performance when there exists a loop in its topological structure. Further, synchronization is also discussed in this work;analysis results illustrate that the proposed model cannot be suppressed to periodic points at sufficiently high coupling strengths. This paper suggests an effective method that may contribute to studying hyperchaos design and coupled chaotic systems.
Micro-expression recognition (MER) remains challenging due to its subtle and fleeting nature. Existing methods often suffer from insufficient training data or rely on handcrafted features. Inspired by recent advanceme...
详细信息
Micro-expression recognition (MER) remains challenging due to its subtle and fleeting nature. Existing methods often suffer from insufficient training data or rely on handcrafted features. Inspired by recent advancements in large language model fine-tuning and visual foundation models (VFMs), we propose HLoRA-MER, a novel framework that combines high-level low-rank adaptation (HLoRA) and a hierarchical fusion module (HFM). HLoRA fine-tunes the high-level layers of a VFM to capture facial muscle movement information, while HFM aggregates inter-frame and spatio-temporal features. Experiments on benchmark datasets demonstrate that HLoRA-MER outperforms state-of-the-art methods, achieving an F1-score of 84.24%\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$84.24\%$$\end{document} and 83.07%\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$83.07\%$$\end{document} on CASME II and SAMM, respectively, with only 197k trainable parameters. Our approach offers a promising solution for MER in both constrained and unconstrained scenarios. The code is available at https://***/CYF-cuber/HLoRA_MER_dinov2.
In multi-modal learning tasks such as video understanding, the most important operations are feature extraction, feature enhancement for single modality and feature aggregation between modalities. In this paper, we pr...
详细信息
In multi-modal learning tasks such as video understanding, the most important operations are feature extraction, feature enhancement for single modality and feature aggregation between modalities. In this paper, we present two attention based algorithms, the Position-embedding Non-local (PE-NL) Network and the Multi-modal Attention (MA) feature aggregation method. Inspired by Non-local Neural Networks and Transformers, our PE-NL is a self-attention liked feature enhancement operation and it can capture long-range dependencies and model relative positions. The MA aggregation method merges visual and audio modals while reduces feature dimension and the number of parameters without losing too much accuracy. Both of PE-NL and MA blocks can be plugged into many multi-modal learning architectures. Our Gated PE-NL-MA network achieves competitive results on Youtube-8M dataset. (c) 2020 Elsevier B.V. All rights reserved.
The timetabling problem (TTP) and vehicle scheduling problem (VSP) are two indispensable problems in public transit planning process. They used to be solved in sequence;hence, optimality of resulting solutions is comp...
详细信息
The timetabling problem (TTP) and vehicle scheduling problem (VSP) are two indispensable problems in public transit planning process. They used to be solved in sequence;hence, optimality of resulting solutions is compromised. To get better results, some integrated approaches emerge to solve the TTP and VSP as an integrated problem. In the existing integrated approaches, the passenger comfort on bus and the uncertainty in the real world are rarely considered. To provide better service for passengers and enhance the robustness of the schedule to be compiled, we study the integrated optimization of TTP and VSP with uncertainty. In this paper, a novel multiobjective optimization approach with the objectives of minimizing the passenger travel cost, the vehicle scheduling cost, and the incompatible trip-link cost is proposed. Meanwhile, a multiobjective hybrid algorithm, which is a combination of the self-adjust genetic algorithm (SGA), large neighborhood search (LNS) algorithm, and Pareto separation operator (PSO), is applied to solve the integrated optimization problem. The experimental results show that the approach outperforms existing approaches in terms of service level and robustness.
This paper investigates the quasi-synchronization of delayed memristive neural networks (MNNs) via a novel hybrid impulsive control algorithm which combines time-triggered and event-triggered impulsive control. The re...
详细信息
This paper investigates the quasi-synchronization of delayed memristive neural networks (MNNs) via a novel hybrid impulsive control algorithm which combines time-triggered and event-triggered impulsive control. The relationship between a predesigned non-negative auxiliary function and a given exponentially decreasing threshold function is used to describe the switching. Under this novel controller, sufficient conditions for the quasi-synchronization are derived by the impulsive differential inequality. In addition, by choosing appropriate parameters or initial conditions such that the initial value of the non-negative auxiliary function is less than that of the event-triggered function, the quasi-synchronization can be realized theoretically as long as the event-triggered impulsive intensity is less than 1. This greatly reduces the conservatism of the existing quasi-synchronization results. Furthermore, the event-triggered rules can avoid the Zeno behavior as long as the event-triggered impulsive intensity is less than 1. This hybrid mechanism can reduce the amount of impulsive control and lessen the network communication. Finally, one example is given to illustrate the validness of the obtained results.
The alternative assessment and decision-making are often complicated and involve enormous decision makers (DMs), which leads to a large group decision making (LGDM) problem, and usually requires to reach consensus wit...
详细信息
The alternative assessment and decision-making are often complicated and involve enormous decision makers (DMs), which leads to a large group decision making (LGDM) problem, and usually requires to reach consensus within a limited time. As a powerful technique in representing linguistic evaluations, the probabilistic linguistic term set is popular to express the opinions of DMs. Under this scenario, DMs who support the same linguistic term are naturally clustered into one subgroup, whereas the distance formulas are inappropriate because whether the DM supports a linguistic term is a dichotomous variable;therefore, Tanimoto coefficient is introduced to calculate the group consensus level. Additionally, an efficient consensus model is proposed to bridge the gap between the subgroup and group opinions. In particular, a weight adjustment function is proposed to process the minority opinions. A novel method including minor and major adjustment manner is proposed to manage the noncooperative behaviors. DMs are allowed to recluster into another subgroup for major adjustment manner;on the contrary, the adjustment direction and quantities are provided for minor adjustment manner. A case study of forest fire emergency decision-making is explored and a simulation is conducted to verify the proposed consensus model in the LGDM problem. The proposed consensus model is suitable for situations where the subgroup and group decision matrices are subject to preference selections and can be applied in other decision problems.
This article is concerned with the optimal tracking performance on linear time-invariant (LTI) discrete-time multi-input-multi-output (MIMO) systems based on quantization, channel noise, and encoding-decoding, as well...
详细信息
This article is concerned with the optimal tracking performance on linear time-invariant (LTI) discrete-time multi-input-multi-output (MIMO) systems based on quantization, channel noise, and encoding-decoding, as well as bandwidth constraints. Meanwhile, power constraint, feedback and feedforward loops constraints of the systems are also taken into account. The tracking performance expression is described by inner-outer factorization, which is obtained by all stabilizing two-degree-of-freedom compensators. The results will illustrate that tracking performance limitation is decided by nonminimum phase zeros and unstable poles of the given plant. Moreover, time delay, encoding-decoding, channel noise, quantization, and bandwidth constraints also impact the performance. The accuracy of the method is verified by simulation examples.
Deblurring aims to restore clear images from blurred ones. Recently deep learning are widely used. Previous methods regard deblurring as dense prediction problems and rarely consider the inverse operation of blur. In ...
详细信息
ISBN:
(纸本)9783030873615;9783030873608
Deblurring aims to restore clear images from blurred ones. Recently deep learning are widely used. Previous methods regard deblurring as dense prediction problems and rarely consider the inverse operation of blur. In this paper, we propose a multi-scale deformable deblurring kernel prediction network for dynamic scene deblurring which uses a coarse-to-fine method to predict the per-pixel deformable deblurring kernel and uses the fusion weight to integrate the latent images in different scales. Since the spatially variable blur scatters pixel information to surrounding sub-pixels and leads to the spatially and quantitively uneven distribution of latent pixel information, the per-pixel deformable deblurring kernel can adaptively select the sub-pixels and linearly combine them into the clean pixel for information aggregation. The multi-scale architecture helps the deformable deblurring kernel enlarge the reception field. The residual image is added to convolution result in each scale to supply refined edges when the kernel cannot cover the areas existing latent pixel information. Besides, we add local similarity loss to constrain deformable deblurring kernel's weight and offset which boosts the deblurring performance. Qualitative and quantitative experiments show that our method can produce competitive deblurring performance.
暂无评论