西南大学

Images taken in low-light conditions are frequently affected by limited visibility, diminished contrast and severe noise, adversely impacting the performance of various computer vision tasks. Most variational-based Retinex decomposition methods mainly depend on integer norms to constrain the illumination and reflectance components. However, this strategy may fail to achieve the ideal Retinex decomposition. In this paper, we propose a Retinex-based variational model that incorporates flexible constraints for both illumination and reflectance. Specifically, we impose the L-p norm constraints with varying values of p to ensure the piece-wise smoothness of the illumination and promote the presence of abundant textures in the reflectance. Moreover, we develop two effective pixel-wise weight matrices that consider variance and gradients of the input image respectively, with the objective of preserving the structural edges of the illumination and retaining more details in the reflectance. In addition, we use an L-p norm to estimate the overall noise level and avoid noise amplification. Through incorporating these constraints, our proposed variational model can obtain a structure-aware illumination and a detail-revealed reflectance. Qualitative and quantitative comparisons on real-world and synthetic datasets indicate that our approach yields results with superior visual quality and outperforms several state-of-the-art algorithms on objective metrics. Besides, our algorithm can also address similar low-level computer vision challenges, such as image dehazing and underwater image enhancement.

This paper focuses on the event-triggered optimal control problem of discrete-time nonlinear multiagent systems (MASs) via the reinforcement learning method. Distinguished from the existing consensus protocols for discrete-time multiagent systems under the ideal communication scenario, the considered model suffers from input saturation, unknown external disturbance, and denial-of-service (DoS) attacks. According to the approximation capability of the radial basis function neural network(RBF NN), the disturbance network is established to tackle the effect of the unknown disturbance on consensus issues. Subsequently, the composite controller is derived with the reinforcement learning strategy to deal with the DoS attacks, where the update frequency of the actor-critic networks and controllers is determined by the novel event-triggered mechanism. Finally, two simulations are formulated to verify the feasibility of the proposed method.

Malicious attacks are a realistic problem that weakens social cooperation. This study investigates malicious strategy attacks with limited strength within the spatial evolutionary snowdrift game. Attackers randomly target cooperators once the cooperation frequency exceeds a threshold. Simulation results reveal that, under the traditional imitation mechanism, malicious attacks suppress cooperation. To address this issue, a weighted memory mechanism and a dynamic recovery security mechanism are proposed. The weighted memory mechanism prioritizes recent interactions by assigning weights to them, enabling rapid responses to emerging threats, while the dynamic recovery mechanism restores cooperation with a defined probability after detecting abnormal attacks. Monte Carlo simulations demonstrate that this security framework effectively mitigates attack damage and promotes cooperation, particularly under high-intensity attacks.

This paper studies the following fractional Schrödinger equation {[ (- Δ )^s u+λ u+ |u|^q-2u =|u|^p-2u in ℝ^N, N≥ 2,; ∫ _ℝ^N|u|^2dx=a>0, ]. where s∈ (0, 1) and 2< p, q ≤ 2_s^*:=2N/N-2s is the fractional critical Sobolev exponent. By giving different assumptions on 2

The inherent properties of Hexagonal Close-Packed (HCP) crystal structure in titanium alloys significantly limits the effectiveness of room-temperature burnishing, thereby restricting performance optimization. By introducing a low-temperature field combined with ball burnishing, the microstructure is optimized and the corrosion and wear properties are improved. TA2 titanium alloy samples were processed via cool-assisted ball burnishing across a temperature range of-110 degrees C to 25 degrees C, followed by systematic electrochemical corrosion and tribological tests. The sample burnished at-110 degrees C exhibited a improvement in surface characteristics, and achieved a 95.32 % reduction in surface roughness, a 45.15 % increase in hardness, and a hardened layer depth of 400 mu m. Electrochemical measurements revealed a positive shift in corrosion potential from-0.683 V to-0.341 V, accompanied by a 85.74 % reduction in corrosion current density and an approximately 12-fold increase in polarization resistance (3.167 x106 Omega & sdot;cm2), confirming its superior corrosion resistance. Under various tribological loads, the-110 degrees C ball burnished sample exhibited the narrowest width and shallowest depth in wear scars, corresponding to a transition from three-body to two-body abrasion and confirming its superior wear resistance. These improvements are attributed to suppressed dynamic recovery and microstructural refinement under cool-assisted field, establishing this method as effective for marine surface enhancement.