Zhaodong Ding

Zhaodong Ding, Zaijiu Shang

In this paper, we study the persistence of invariant tori of integrable Hamiltonian systems satisfying Rüssmann's non-degeneracy condition when symplectic integrators are applied to them. Meanwhile, we give an estimate of the measure of the set occupied by the invariant tori in the phase space. On an invariant torus, the one-step map of the scheme is conjugate to a one parameter family of linear rotations with a step size dependent frequency vector in terms of iteration. These results are a generalization of Shang's theorems (1999, 2000), where the non-degeneracy condition is assumed in the sense of Kolmogorov. In comparison, Rüssmann's condition is the weakest non-degeneracy condition for the persistence of invariant tori in Hamiltonian systems. These results provide new insight into the nonlinear stability of symplectic integrators.

Zhaodong Ding, Zaijiu Shang, Bo Xie

We prove a Nekhoroshev-type theorem for nearly integrable symplectic map. As an application of the theorem, we obtain the exponential stability symplectic algorithms. Meanwhile, we can get the bounds for the perturbation, the variation of the action variables, and the exponential time respectively. These results provide a new insight into the nonlinear stability analysis of symplectic algorithms. Combined with our previous results on the numerical KAM theorem for symplectic algorithms (2018), we give a more complete characterization on the complex nonlinear dynamical behavior of symplectic algorithms.

Chenglong Li, Tao Wang, Zhaodong Ding et al.

RGBT tracking usually suffers from various challenging factors of low resolution, similar appearance, extreme illumination, thermal crossover and occlusion, to name a few. Existing works often study complex fusion models to handle challenging scenarios, but can not well adapt to various challenges, which might limit tracking performance. To handle this problem, we propose a novel Dynamic Disentangled Fusion Network called DDFNet, which disentangles the fusion process into several dynamic fusion models via the challenge attributes to adapt to various challenging scenarios, for robust RGBT tracking. In particular, we design six attribute-based fusion models to integrate RGB and thermal features under the six challenging scenarios respectively.Since each fusion model is to deal with the corresponding challenges, such disentangled fusion scheme could increase the fusion capacity without the dependence on large-scale training data. Considering that every challenging scenario also has different levels of difficulty, we propose to optimize the combination of multiple fusion units to form each attribute-based fusion model in a dynamic manner, which could well adapt to the difficulty of the corresponding challenging scenario. To address the issue that which fusion models should be activated in the tracking process, we design an adaptive aggregation fusion module to integrate all features from attribute-based fusion models in an adaptive manner with a three-stage training algorithm. In addition, we design an enhancement fusion module to further strengthen the aggregated feature and modality-specific features. Experimental results on benchmark datasets demonstrate the effectiveness of our DDFNet against other state-of-the-art methods.