The existing results of passivity for neural networks mainly concentrated on first-order derivative of the states, whereas it is also significant to study the passivity with high-order derivative. In this paper, a class of memristor-based inertial neural networks (MINNs) with external inputs and outputs are concerned. First, by choosing appropriate variable transformation, the original networks are rewritten as first-order differential equations. Then a criterion of passivity for the MINNs is presented by nonsmooth analysis and linear matrix inequality (LMI) techniques, which could also result in NP-hard problem since it needs at least exponential-time to solve the passivity condition. Meanwhile, based on the obtained passivity criterion, asymptotic stability criterion is accordingly derived for MINNs. In order to avoid the NP-hard problem, we employ matrix-analysis-techniques with the property that the difference-matrix between the given matrix and the proposed matrix is seminegative definite. Then the time complexity for solving the proposed passivity criterion is reduced to constant-time with respect to the number of LMIs. Robust passivity for MINNs is also studied for bounded uncertain parameters. The MINN widens the application ranges for designing neural networks. Finally, relevant simulation examples are given to show the effectiveness of the obtained results.

Without confines of the continuous-time domain, this paper addresses synchronization control problem of neural networks in the face of multiple proportional delays on general time scales. The idea to deal with proportional delays is to propose a class of -type allowable time scales on which we design an appropriate controller to achieve exponential synchronization based on a calculus theory on time scales and Lyapunov function/functional methods. It is shown that adopting properties of -type time scales is an effective approach to establish synchronization for the networks with proportional delays. This helps us to have insight into the synchronization problems on general intermittent time domain. Finally, simulation examples are given to illustrate the effectiveness of the theoretical results.

Current background subtraction algorithms are sensitive to sudden changes. In this paper, we propose a multi-feature fusion scheme to background subtraction for video sequences with strong background changes. We reconstruct the whole videos frame by frame by fusing several video features. In this fusing step, we design an energy function based on enforcing every features with an equal weight. By comparing reconstruction videos with the original videos, pixels with small differences are classified as background pixels. Thus, we can identify background areas in advance and then we construct a contour-based mask combining mechanism. Experimental results conducted on the OTCBVS, BMC 2012 and PETS 2001 datasets show that our method improves the performance of the Zivkovic's GMM and SubSENSE for video sequences with strong background changes.

The existing results for multiperiodicity of threshold-linear networks (TLNs) are scale-free on time evolution and hence exhibit some restrictions. Due to the nature of the scale-limited activating set, it is interesting to study the dynamical properties of neurons on time scales. In this paper we analyze and obtain results concerning nondivergence, attractivity, and multiperiodic dynamics of TLNs on time scales. Using the notion of exponential functions on time scales, we obtain results for scale-limited type criteria for boundedness and global attractivity of TLNs. Moreover, by constructing simple algebraic inequalities over scale-limited activating sets, we achieve results regarding multiperiodicity of TLNs. This will show that each scale-limited activating set depends on scale-synchronous self-excitation, and the existence of inactive neurons will slow down convergence of TLNs. At the end of the paper, we perform computer simulations to illustrate the obtained new theories.

The classic Gaussian mixture model is based on the statistical information of every pixel; it is not robust to light changes. Before analysing every pixel in videos, it must be decoded to raw videos. In this letter, the method combining video coding and the Gaussian mixture model together is proposed. We use intra mode and motion vectors to find the foreground macroblock, then add one overhead flag in the compressed video to indicate it. In the decoder, we just decode possible foreground areas and detect moving objects in these areas. In our experiments, we test this method on two datasets, both of them with unique, dynamic, illumination conditions. Results show that the proposed method is effective to detect moving objects and easily assemble to current automated video surveillance systems.

Motion saliency is the key component for the video saliency model, and attracts great research interest. However, there is few universal predictor of motion saliency. In this paper, a novel method for generation of motion saliency is proposed, in which motion saliency map is obtained through the multi reference frames, and enhanced by spatial saliency information. The proposal can obtain more detailed information about motion feature and extract the salient object more integrally. The experiment results shows that our proposal can achieve 95.4% of the ROC area of a human based control in motion channel, whereas the classical Itti's model achieves 78.5%.

The existing approaches to the multistability and multiperiodicity of neural networks rely on the strictly excitatory self-interactions of neurons or require constant interconnection weights. For periodically oscillated discrete-time neural networks (DTNNs), it is difficult to discuss multistable dynamics when the connection weights are periodically oscillated around zero. By using transient excitatory self-interactions of neurons and sigmoidal nonlinearities, we develop an approach to investigate multiperiodicity and attractivity of periodically oscillated DTNNs with time-varying and distributed delays. It shows that, under some new criteria, there exist multiplicity results of periodic solutions which are locally or globally exponentially stable. Computer numerical simulations are performed to illustrate the new theories.